Microsoft Entra sign-in logs help you answer a practical question: why did an access attempt succeed or fail? They show the identity, application, target resource, client, location signals, authentication details, and policy results associated with a sign-in request. This guide uses the Microsoft Entra admin center to investigate failed sign-ins without assuming that every unfamiliar location or error means an account compromise.
Prerequisites and scope
- Use an account with at least the Reports Reader role to view sign-in logs.
- Use the Microsoft Entra admin center at entra.microsoft.com.
- Understand that sign-in logs are system-generated records. They cannot be edited or deleted by an administrator.
- Make sure the date range and retention available to your tenant are sufficient for the incident. Retention and advanced analysis options depend on licensing and configuration.
Do not block on location alone. Mobile carriers, VPNs, proxies, cloud services, and corporate egress points can make a legitimate sign-in appear to originate from an unexpected city or country.
1. Open the sign-in logs
- Sign in to the Microsoft Entra admin center.
- Open Entra ID and select Monitoring & health.
- Select Sign-in logs.
The legacy sign-in view focuses on interactive user sign-ins. Depending on your tenant and the current portal experience, you may also see non-interactive user, service principal, managed identity, or agent-related log types. Select the log type that matches the event you are investigating.
2. Start with a narrow filter
Use a small time window and one known signal rather than searching the entire tenant first. Useful filters include:
- Date: Start with the time reported by the user, then account for time-zone differences.
- User: Filter to the affected account when the incident concerns one person.
- Status: Select failure or success as appropriate. A successful sign-in following repeated failures may be the most important sequence.
- Application: Narrow to Exchange Online, SharePoint Online, Teams, a line-of-business app, or another named application.
- IP address, resource, device, or location: Use these to separate a known device or service from an unusual source.
- Conditional Access: Filter or review policy results when a policy may have blocked, interrupted, or required additional authentication.
Apply the filter and sort by time. Record the filter values and the time zone in your investigation notes so another administrator can reproduce the search.
3. Open the event details
Select a sign-in event and review the detail panes systematically:
- Basic information: Confirm the user, timestamp, application, resource, and sign-in type.
- Status: Record the error code and failure reason. Copy the code exactly before searching for an explanation.
- Device info: Compare the operating system, browser, device identifier, and managed or compliant state with the user’s known device.
- Location: Treat the displayed location as a network-derived signal, not a physical proof of where the person was.
- Authentication details: Check whether a password, MFA, authentication-strength requirement, or another method succeeded or failed.
- Conditional Access: Identify which policies applied and whether the result was success, failure, or not applied. The policy result explains enforcement, not necessarily the root cause of the user’s password or device problem.
4. Correlate the event before taking action
Build a short timeline. Compare failed sign-ins with:
- The user’s reported time, device, and network.
- Other sign-ins for the same account and application.
- Microsoft Entra audit activity, Microsoft 365 audit activity, and endpoint security alerts.
- Recent password changes, MFA registration changes, risky-user or risky-sign-in detections, and help-desk reports.
Repeated failures from one known device may indicate a stale password, cached credentials, time drift, or an authentication policy issue. A successful sign-in from an unfamiliar device followed by mailbox or file activity deserves a stronger incident response, including account protection and evidence preservation.
5. Use the error code carefully
Do not translate an error code into a diagnosis without reading the event context. The same broad failure category can result from Conditional Access, MFA, device compliance, application configuration, credentials, or a service dependency. Search the exact code in Microsoft’s Microsoft identity platform error-code reference, then compare the explanation with the event’s policy and authentication details.
Verification checklist
- Can you reproduce the failure with the same user, application, device state, and network without weakening security controls?
- Does the sign-in detail show the expected policy and authentication method?
- After remediation, does a new sign-in event show success and the expected device and location signals?
- Did the change create a new risk, such as bypassing MFA or excluding too many users from Conditional Access?
Common limitations
Sign-in data is not always available immediately, and the portal’s visible history and analysis features vary with tenant configuration and licensing. The logs show the request received by Microsoft Entra; they do not prove that a person physically controlled the device or that an IP address belongs to a specific individual. For long-term monitoring, export or route the appropriate logs to a supported analysis platform instead of relying on a manual portal search.
Related Guava guidance
- How to Set Up Multi-Factor Authentication with Microsoft Authenticator
- How to Protect Microsoft 365 Administrators With Phishing-Resistant MFA

