SEND Exam Guide: Skills, Study Decisions, and a Practical Preparation Roadmap
The available catalogue label identifies this assessment as SEND, but the supplied official research does not provide an exam page, blueprint, score, duration, language list, prerequisites, or delivery rules. The evidence instead centers on Microsoft messaging, SMTP submission, Exchange mail flow, Microsoft Graph, nondelivery reports, and Windows networking. This guide helps candidates decide whether that technical scope matches their goal, which subjects to study first, how to build useful practice, and which exam details must be confirmed before scheduling.
What does the SEND exam appear to validate?
The supplied evidence supports a technical preparation scope around sending and troubleshooting email rather than a verified official exam blueprint. Candidates should treat Microsoft Graph message operations, Exchange routing, SMTP connection methods, delivery failures, and supporting network behavior as study areas suggested by the research—not as confirmed exam domains.
Microsoft Graph represents email as a message resource. A candidate working in this area should understand the difference between creating and sending a message in one action and creating a draft, modifying it, and sending it later. The official Graph guidance also explains that a message ID can change when a message is copied or moved, so further processing must use the message's current ID. See https://learn.microsoft.com/en-us/graph/outlook-create-send-messages.
The Graph send process is not the same as final delivery. The supplied research describes draft creation, notification of the transport service, movement into the transport pipeline, policy evaluation, routing, recipient delivery, and report generation as separate stages. A sound study plan therefore needs both API knowledge and mail-flow reasoning.
Do not describe these subjects as official measured domains unless an authoritative SEND exam page confirms them. For this article, they form an evidence-led working map that can guide preparation while the candidate verifies the actual exam specification.
Who should use this preparation plan?
This plan best fits a candidate who administers, develops, integrates, or troubleshoots Microsoft-based email flows. It is especially relevant when the work involves Microsoft Graph message automation, Exchange Server Send connectors, Microsoft 365 device or application submission, SMTP errors, or network paths that affect RPC or socket communication.
An application developer should begin with message resources, draft and send behavior, sender permissions, attachments, and the distinction between an accepted API request and successful delivery. An Exchange administrator should begin with connector selection, address spaces, routing, source servers, smart hosts, DNS, and TLS. A support engineer should prioritize NDR interpretation and a repeatable diagnostic sequence.
The research also includes the Microsoft Release Communications MCP Server. That material is relevant to a candidate who needs to query Microsoft 365 Roadmap or Azure Updates through an MCP-compatible client, but it should not displace core mail-flow study unless the official exam outline explicitly includes it. The source describes an endpoint, client configuration, filtering, and retrieval of release information; it does not connect those capabilities to a verified SEND exam objective.
Use your current job tasks to choose emphasis. If you cannot explain where a message is created, how it is authenticated, which service routes it, and why a failure report was produced, begin with fundamentals rather than memorizing isolated error codes.
Which skills should you measure before studying?
Measure your ability to explain a complete message path and diagnose the first failing boundary. A useful self-assessment covers Graph operations, SMTP submission choices, Exchange connector configuration, TLS and identity, NDR interpretation, and network connectivity. These are practical capability checks derived from the supplied sources, not confirmed percentage-weighted exam domains.
For Graph, test whether you can explain when a message is a draft, when it is sent, where each state is normally stored, and why an ID must be refreshed after a move or copy. You should also be able to reason about from and sender properties, including the difference between send-as rights and delegated send-on-behalf behavior.
For mail submission, test whether you can select among client SMTP submission, SMTP relay, Direct Send, and High Volume Email based on recipient scope, authentication, TLS, mailbox requirements, and whether Internet relay is needed. The Microsoft 365 device and application guidance provides the comparison; use it as a decision table rather than learning each method in isolation.
For Exchange Server, test whether you can identify the role of a Send connector, its address space, network routing choice, scope, and source servers. The official connector documentation states that connectors handle outbound SMTP connections from source Exchange servers to destination email servers and that external mail flow requires a configured connector or an Edge Transport subscription.
For troubleshooting, test whether you can classify an NDR as an address, authentication, authorization, certificate, routing, policy, rate, or remote-server problem. Then test whether you can name the next diagnostic action without guessing. The NDR documentation supplies error descriptions and administrator details that support this classification.
What exam information is not evidenced?
No supplied official source confirms the SEND exam's question count, duration, scoring method, passing score, languages, prerequisites, registration process, price, delivery method, testing locations, retirement status, or accommodations. Do not rely on a third-party listing for these details without checking the issuing organization's current exam page.
The catalogue identifier is SEND (1:exam:1631:ExamArticle), but that identifier does not establish a vendor, certification path, or official blueprint. The Microsoft pages supplied here are technical product documentation, not an exam announcement or candidate handbook. They can support subject preparation, but they cannot verify scheduling rules.
Before paying or booking, confirm the exam's exact title, issuing organization, current status, official objectives, eligibility requirements, available delivery options, identity rules, rescheduling policy, and result reporting process from the authoritative registration source. Record the page and the date you checked it because these details can change.
If the official exam page names domains that differ from the technical scope in this guide, follow the official blueprint. Use the source-grounded material here as supporting study content only. This distinction prevents a candidate from spending preparation time on a plausible topic that is not actually assessed.
How should you study Microsoft Graph message sending?
Start with the message lifecycle, then practice the API choices that implement it. Learn how a single sendMail action differs from a draft workflow, how message identifiers behave after movement, and how the service processes a request after the API response. This sequence builds operational understanding instead of encouraging endpoint memorization.
The supplied Graph material says that emails are represented by the message resource. A message may be created and sent in one action, or created as a draft and sent later. Draft messages are normally saved in Drafts, while sent messages are normally saved in Sent Items. A candidate should be able to predict the state after each operation and identify what must be queried next.
The send-process documentation explains that an HTTP response of 202 Accepted indicates successful handling of the initial method request, not proof that every recipient received the message. After the response, transport can still copy the message into its pipeline, apply policy, route it, deliver it, or generate a report. Make this distinction central to your notes.
Study delegated sending carefully. The from property can be changed when an administrator assigns send-as rights, while the sender property can represent a delegate sending on behalf of a mailbox owner. These are authorization concepts, not interchangeable fields. Build a comparison table with the identity displayed to recipients, the permission needed, and the API property involved.
A productive practice exercise is to design two workflows on paper: an immediate send and a draft-review-send process. For each, list the request, expected message state, identifier used for the next operation, possible authorization failure, and evidence that would establish whether transport—not merely submission—succeeded.
How should you learn SMTP submission choices?
Choose the sending method from the business requirement first. Client SMTP submission uses a cloud mailbox and authentication; SMTP relay uses an inbound connector and a static IP or certificate-based identity; Direct Send is unauthenticated and limited to Microsoft 365 or Office 365 recipients; High Volume Email is intended for large-volume internal sending. Verify the current product rules before implementation.
Client SMTP submission is the natural study case for a device or application that sends through a cloud mailbox. The supplied comparison specifies TCP port 587 or 25, TLS support, and Microsoft 365 or Office 365 credentials. It also states that SMTP AUTH is disabled for organizations created after January 2020 but can be enabled per mailbox. Treat authentication policy as an administrative dependency, not a coding detail.
SMTP relay is a different trust model. The documentation describes an inbound connector that authenticates the device or application, commonly using a TLS certificate or public IP address. It does not require a licensed cloud mailbox, but the sending address must be associated with an accepted domain. A static IP must be visible to Microsoft 365 or Office 365 when IP-based authentication is used.
Direct Send sends unauthenticated mail as an external email server directly to Microsoft 365 or Office 365. It uses TCP port 25 and is intended for recipients in the organization's accepted domains rather than Internet relay. Because some firewalls or ISPs block port 25, connectivity must be checked before choosing this method.
High Volume Email has a distinct purpose: sending large volumes to internal recipients with an HVE account authenticated through credentials or OAuth. Do not use the name as a general solution for external bulk mail. The source comparison states that the method does not relay to the Internet through Microsoft 365 or Office 365.
Make a four-column decision sheet with method, recipient scope, authentication, and infrastructure requirement. Fill it from the official comparison, then add a fifth column for the failure you would expect if that requirement were missing. This turns product documentation into scenario practice.
What Exchange Send connector concepts matter most?
Study Send connectors as routing decisions, not as merely outbound SMTP settings. You should be able to connect a destination domain to an address space, explain whether DNS or a smart host is used, identify the source servers, and understand connector scope. These concepts provide the framework for analyzing why a message takes one outbound route rather than another.
Microsoft's Exchange Server documentation says that a Send connector is selected during message categorization and is used for outbound SMTP connections from source Exchange servers to destination email servers. Important settings include usage type, network settings, address spaces, scope, and source servers. Turn each setting into a question: what traffic does it match, where does it go, and which server sends it?
External mail flow is not automatically enabled by installing Exchange. The supplied guidance states that no Send connectors exist by default for external mail flow and that an organization must create one or subscribe an Edge Transport server. Internal Exchange communication is different because implicit, invisible connectors handle intra-organization transport.
Do not overlearn obsolete parameters. The research states that IsCoexistenceConnector and LinkedReceiveConnector are no longer available. If a practice resource asks you to configure either parameter, check its product version and credibility before accepting the premise.
Message size is another area where context matters. The connector documentation states that the default maximum message size is increased to 35 MB, approximately 25 MB because of Base64 encoding. Keep that fact attached to Exchange connector behavior and do not generalize it as a universal Microsoft 365 attachment limit.
For practice, draw a route for an internal recipient, an Internet recipient resolved through DNS, and a destination reached through a smart host. Mark the connector's address space and source server at each decision point. Then alter one setting and explain the resulting routing consequence.
How do you troubleshoot an NDR without guessing?
Read the complete nondelivery report before changing configuration. Start with the enhanced status code, identify whether the failure is temporary or permanent, inspect the stated cause, and separate the submitting service from the receiving organization. Then choose one targeted check. An NDR is diagnostic evidence; treating every bounce as an Exchange configuration problem wastes time.
The official NDR guidance explains that a delivery status notification typically includes an error code, a reason the message was not delivered, possible solutions, additional help, and technical administrator details. It also warns that the Microsoft 365 logo does not necessarily mean Microsoft 365 caused the error. Identify the generating server and recipient-side system before assigning ownership.
For address failures, compare the recipient address with directory or domain information. Examples in the supplied facts include 5.1.10 Recipient not found and 5.4.1 Recipient address rejected: Access denied. These indicate different investigation paths from an authentication failure. Do not fix a connector when the stated problem is simply that the recipient does not exist.
For authentication and authorization, distinguish 5.7.1 Client was not authenticated, 5.7.57 Client was not authenticated to send anonymous mail during MAIL FROM, and 5.7.60 SMTP; Client doesn't have permissions to send as this sender. The first two point toward authentication or relay configuration; the latter points toward sender permission and identity alignment.
For certificate and TLS failures, check certificate validity, subject or SAN matching, trust chain, and whether the destination supports TLS. The research includes 4.7.321 starttls-not-supported, 4.7.322 certificate-expired, 4.7.325 certificate-host-mismatch, and 450 4.4.317 UntrustedRoot. Learn the evidence each code implies instead of treating all certificate errors as one category.
For policy, reputation, and limits, investigate the exact message. The research includes 5.7.606-649 for a banned sending IP, 5.7.800 for a banned sender domain, 5.7.367 for SPF or DKIM failures in forwarded or relayed mail, and 5.2.2 for a submission quota being exceeded. These are not solved by repeatedly resending the same message.
Practice with a five-step worksheet: code, responsible boundary, likely cause, confirming evidence, and safe next action. Include the full diagnostic text when available. This habit is more transferable than memorizing a long list of codes without their context.
What network foundations should you revise?
Revise network behavior only to the depth needed to explain a failure. For RPC, understand the initial connection to TCP port 135, endpoint mapper negotiation, dynamic ports, firewall requirements, and diagnostic tools. For Winsock, understand connected sockets, partial sends, blocking behavior, and error handling. These subjects support troubleshooting; they should not become disconnected theory.
Microsoft's RPC guidance describes the endpoint mapper as the service that helps clients locate server applications using port and UUID information. The client initially contacts TCP port 135, negotiates a dynamic port through the endpoint mapper, and then connects to that dynamic port. A firewall must allow both port 135 and the assigned dynamic ports.
A common mistake is opening only port 135 and concluding that RPC is available. The source identifies the dynamic port as the later communication path and describes PortQry as a way to test connectivity. It also recommends network traces when the basic test does not reveal the cause. Use this order: reachability, endpoint negotiation, dynamic-port access, then packet-level evidence.
For the Winsock send function, remember that the return value is the number of bytes sent and can be less than the requested length. On a nonblocking stream-oriented socket, the operation can write between 1 and the requested length depending on buffer availability. Application code must account for partial transmission rather than assuming one call sends everything.
The source also states that send can block when no buffer space is available unless the socket is nonblocking, and that concurrent send calls on the same stream-oriented socket can cause unintended interleaving. These details are useful when diagnosing an application that appears to submit incomplete or corrupted data.
Create one page that maps symptom to layer: DNS or name resolution, TCP connection, TLS negotiation, authentication, SMTP authorization, transport policy, or recipient acceptance. For each layer, name one command, log, or response that would provide evidence. Avoid changing several layers at once.
How can you build a safe practice lab?
Use a controlled tenant, test mailbox, or documented simulation rather than production mail flow. The goal is to reproduce decision points—draft versus sent state, authentication choice, connector route, TLS identity, and NDR interpretation—without sending unsolicited messages or attempting to bypass protection. Practice configuration reasoning even when live access is unavailable.
For a Graph exercise, create a message model with recipients and attachments, then trace the draft and send paths. Confirm that subsequent operations reference the current message ID. Record what 202 Accepted tells you and what it does not tell you. Use test recipients under your control and preserve request and response information for review.
For SMTP exercises, write a requirements brief before touching settings. State whether recipients are internal or external, whether a licensed mailbox is available, whether the device supports TLS, whether a static IP or certificate can be supplied, and whether port 25 or port 587 is reachable. Select the method only after those answers are explicit.
For Exchange exercises, create a route diagram rather than relying on a live change. Identify address space, DNS or smart-host routing, scope, and source servers. If you have an authorized lab, use message tracing and connector logs to compare the expected route with the actual route. Never test relay behavior against third-party systems without permission.
For NDR practice, use the documented codes as case prompts. Ask what the code says, whether the issue is on the sending or receiving side, what configuration could produce it, and what evidence would confirm that theory. Do not seek leaked questions or use exam dumps as a substitute for understanding; they cannot establish the live exam's content or guarantee a pass.
Which study sequence gives the best return?
Study in dependency order: message lifecycle, submission method selection, Exchange routing, failure interpretation, and network diagnostics. This order prevents a common error—trying to learn error codes before understanding where a message travels. Revisit each earlier topic through a later scenario so that the knowledge becomes connected rather than a collection of notes.
First, learn the vocabulary and lifecycle. Define message resource, draft, Sent Items, transport pipeline, connector, address space, accepted domain, smart host, endpoint mapper, NDR, and enhanced status code. Write a short explanation of the complete path from application request to recipient or bounce.
Next, compare the four Microsoft 365 sending methods. Build a matrix from the official source and mark recipient scope, authentication, TLS, mailbox dependency, IP or certificate requirement, and whether the method relays to the Internet. Explain why a method that works for internal devices may fail for external recipients.
Then study Exchange routing. Practice selecting a Send connector during categorization, matching address spaces, and deciding between DNS and smart-host delivery. Add message-size and connector-version notes only after the core routing model is clear.
After that, work through NDR families. Start with recipient and authentication errors, then TLS and certificate errors, followed by routing, reputation, and rate-limit cases. For each case, write a first diagnostic action and a configuration change you would not make until evidence supported it.
Finish with RPC and Winsock fundamentals. These are most useful when a case requires you to distinguish application behavior from network reachability. Test yourself by explaining why port 135 alone may not be enough for RPC and why a socket send may require handling fewer bytes than requested.
At the end of each session, close the documentation and reconstruct the decision from memory. Reopen the source only to correct the specific gap. This method reveals whether you understand a rule or merely recognize its wording.
What should a four-stage roadmap look like?
A practical roadmap has four stages: establish scope, build concepts, troubleshoot scenarios, and verify readiness. The duration should match your background and the official exam date, which are not supplied here. Do not create a calendar around an assumed question count or passing score; use demonstrated capability and the verified blueprint instead.
Stage one is scope control. Obtain the official SEND objectives and mark every objective as known, partly known, or unknown. Compare those objectives with the evidence-led areas in this guide. Remove topics that the official outline excludes, and add official topics that the supplied research does not cover. This prevents catalogue context from becoming an accidental substitute for the blueprint.
Stage two is concept construction. Produce four artifacts: a Graph lifecycle diagram, a submission-method decision matrix, an Exchange connector route map, and an NDR classification sheet. Each artifact should contain source links and a short explanation in your own words. If you cannot explain a decision without copying the documentation, the topic needs another pass.
Stage three is scenario troubleshooting. Work from symptoms to evidence: an accepted API request followed by nondelivery, a scanner that cannot authenticate, a certificate mismatch, a relay denial, an RPC server unavailable message, and a partial socket send. For every scenario, state the failing layer and the next check before proposing a fix.
Stage four is readiness verification. Use mixed, closed-book scenarios and review errors by category. A weak area is not simply a missed fact; it may be a failure to identify the responsible boundary, interpret a status code, or choose an appropriate diagnostic action. Re-test the category after correcting the underlying model.
When you have a scheduled date, reserve the final review for official objectives, configuration distinctions, and your error log. Avoid introducing new unofficial material immediately before the appointment. Confirm current delivery and identification requirements through the official registration source rather than through a static study article.
What mistakes undermine SEND preparation?
The most damaging mistakes are scope confusion, false certainty about delivery, and memorization without diagnosis. Candidates should not treat Microsoft product documentation as proof of an exam blueprint, assume 202 Accepted means delivery, select SMTP methods by habit, or memorize NDR codes without identifying the failing boundary.
Do not mix Microsoft Graph and SMTP concepts casually. Graph is an API model and workflow; SMTP submission is a protocol and service configuration. They can participate in the same application design, but their authentication, state handling, and diagnostic evidence differ. Keep separate notes, then add a final section explaining how the layers interact.
Do not confuse SMTP relay with Direct Send. Relay uses an inbound connector and an authenticating identity such as a static IP or certificate. Direct Send is unauthenticated and does not relay through Microsoft 365 or Office 365 to Internet recipients. A scenario's recipient scope should drive the choice.
Do not treat every 5.x.x code as a permanent sender-side fault or every 4.x.x code as harmless. Read the description and diagnostic detail. A certificate, recipient, policy, authentication, routing, or remote-server issue demands a different response. Re-sending without correcting the cause can obscure the pattern and may worsen rate or reputation problems.
Do not copy outdated Exchange configuration advice. The supplied connector source explicitly notes parameters that are no longer available, while the SMTP guidance includes product-specific authentication and port requirements. Check the version and last-updated context of any secondary resource.
Finally, do not use exam dumps, leaked questions, or answer memorization. Such material is not a reliable representation of the current assessment, does not build operational skill, and cannot guarantee passing. Use official objectives and documentation, then practice explaining decisions in unfamiliar scenarios.
How should you decide whether to schedule?
Schedule only after confirming the official exam identity and registration details and after demonstrating that you can solve mixed technical scenarios without relying on answer recall. The supplied research does not establish a score threshold or exam format, so readiness must be based on the verified blueprint, documented requirements, and repeatable reasoning.
Before scheduling, check the issuing organization's current page for the exam title, code, prerequisites, delivery options, language availability, appointment rules, identification requirements, rescheduling terms, and result process. None of those details is evidenced by the supplied sources. Save the official link you used and recheck it close to booking.
For technical readiness, select a random scenario from each study area. Explain a Graph state transition, choose a sending method, trace a connector route, classify an NDR, and diagnose a network symptom. If your answer depends on a remembered phrase but you cannot identify the evidence or next action, continue studying.
For administrative readiness, make sure your planned environment and identity documents meet the official provider's requirements. Do not infer whether the exam is online, at a test center, proctored, or otherwise delivered from the Microsoft technical pages. Delivery details belong to the official exam and registration documentation.
The final decision should be based on evidence you can verify: current objectives, completed practice artifacts, corrected error categories, and confirmed booking rules. If the official outline is unavailable or materially different from the catalogue label, pause and resolve that uncertainty before committing money or time.
What should you do next?
Begin by locating the official SEND exam page and copying its current objectives into a study checklist. Then use the Microsoft sources below to fill the relevant technical gaps, starting with the message lifecycle and submission-method comparison. Your next useful action is not to collect more questions; it is to identify which requirement, route, or diagnostic step you cannot yet explain.
Create the Graph lifecycle diagram and the four-method SMTP matrix first. Next, draw one Exchange route and complete five NDR worksheets using different failure categories. Finish with an RPC connectivity path and a Winsock partial-send explanation. Keep every conclusion tied to a source or clearly labeled as a preparation recommendation.
Return to the official registration source before scheduling. Confirm all time-sensitive or administrative facts there, because they are not present in the supplied evidence. This workflow gives you a defensible preparation scope while avoiding unsupported claims about the exam itself.
Conclusion
The evidence supports preparation for Microsoft messaging and delivery scenarios, but it does not verify the SEND exam's official blueprint or administrative format. Use the technical sources to build connected skills: understand Graph state and identity, select SMTP methods from requirements, trace Exchange routes, interpret NDRs, and separate network faults from mail-flow faults. Verify the exam provider's current objectives and scheduling rules before booking, and use scenario-based reasoning rather than dumps or memorized answers.
