Incident response plan testing against NIST SP 800-61 Rev. 3 means exercising your plan across the six NIST CSF 2.0 Functions that replaced the old four-phase lifecycle in April 2025: Govern, Identify, Protect, Detect, Respond and Recover. Run a tabletop scenario aimed at each Function, then measure detection, decision and recovery times against targets set beforehand.
Most incident response plans were written to the 2012 version of NIST’s guidance, and most incident response plan testing amounts to whatever scenario felt urgent that quarter. This guide covers what changed in Rev. 3, which tabletop scenario tests each part of the framework, and what to measure so you can show the plan works. It’s for CISOs, IT directors and the security and compliance leads who own the plan.
What are the key phases of incident response under NIST SP 800-61 Rev. 3?
NIST SP 800-61 Rev. 3, Incident Response Recommendations and Considerations for Cybersecurity Risk Management: A CSF 2.0 Community Profile, was published on April 3, 2025. It supersedes Rev. 2, the Computer Security Incident Handling Guide (2012).
You can read the full publication on NIST’s SP 800-61 Rev. 3 page. It’s built on the NIST Cybersecurity Framework (CSF) 2.0, so the two are easiest to use together.
Rev. 2 described a four-phase lifecycle:
- Preparation
- Detection and Analysis
- Containment, Eradication and Recovery
- Post-Incident Activity
Rev. 3 drops that cycle and organizes incident response around the six CSF 2.0 Functions instead. They fall into two groups:
- Preparation and lessons learned: Govern, Identify and Protect. This is the work you do before an incident and the improvement work you do after one.
- Incident response: Detect, Respond and Recover. This is the work you do during an incident.
Rev. 3 also treats incident response as ongoing. Your organization prepares, learns and improves whether or not an incident is active, so lessons learned sit under Identify’s Improvement category instead of a final phase.
How do the Rev. 2 phases map to the Rev. 3 Functions?
| Rev. 2 phase | Rev. 3 CSF 2.0 Function(s) | What moved or changed |
|---|---|---|
| Preparation | Govern, Identify, Protect | Preparation now includes governance: roles, decision authority, risk strategy and supply chain. |
| Detection and Analysis | Detect (continuous monitoring, adverse event analysis), Respond (incident analysis) | Detection is ongoing monitoring. Analysis of a declared incident sits under Respond. |
| Containment, Eradication and Recovery | Respond (incident management, mitigation, reporting and communication), Recover (recovery plan execution, recovery communication) | Recovery is its own Function, with its own communication requirements. |
| Post-Incident Activity | Identify (Improvement) | Lessons learned feed back into risk management continuously, not just at the end. |
Does a plan written to Rev. 2 still work?
Mostly. The response steps are still sound. The usual gaps are governance and improvement: who has authority to declare an incident and make high-impact decisions, how incident response ties into enterprise risk management, and whether lessons learned actually get closed out. Those are also the gaps a phase-by-phase tabletop program is most likely to surface.
How does incident response plan testing work for each phase?
A tabletop exercise is a facilitated, discussion-based walkthrough of a realistic incident. (For the full structure, see what a good tabletop exercise includes.) To test against Rev. 3, aim each scenario at one Function. That way every exercise has a clear question to answer, and across a year you cover the whole framework.
Govern: who has the authority to make the call?
What it covers: roles and responsibilities, decision authority, risk tolerance, policy, and oversight of third parties.
Scenario: At 2 a.m. on a holiday weekend, the on-call engineer finds ransomware notes on two file servers. The CISO is on a flight, and the threat actor has posted a 72-hour deadline.
Injects to escalate:
- The cyber insurer’s breach hotline says to engage only panel firms.
- A board member hears about it and emails the CEO directly.
- Legal asks whether this could be material for disclosure purposes. (For public companies, see the SEC materiality decision points your plan needs.)
What to observe: whether a named deputy can formally declare an incident, who approves bringing in outside counsel and the IR firm, and whether anyone knows the insurer’s notification requirements.
Measure: time from first report to formal incident declaration, and the number of decisions that stalled because no one was named as the owner.
Identify: do you know what was hit and why it matters?
What it covers: asset inventory, data classification, risk assessment, and improvement (where Rev. 3 now puts lessons learned).
Scenario: An external researcher reports an exposed database. It belongs to a development server no one remembers deploying, and it seems to hold a copy of production customer data.
What to observe: how long it takes to find an owner, whether the data’s sensitivity can be confirmed, and whether this exposure was already flagged in a past assessment or after-action report.
Measure: time to identify the asset owner and data classification, plus the share of action items from the last exercise that are actually closed.
Protect: were the safeguards in place before you needed them?
What it covers: identity and access management, awareness, data security, platform security, and infrastructure resilience.
Scenario: A privileged administrator approves an MFA push they didn’t request after a run of prompts at night. (See how MFA fatigue attacks work.) The attacker registers a new device and creates a forwarding rule.
What to observe: whether the logs you need were enabled and kept long enough, how fast sessions and tokens can be revoked, and whether backups are isolated from the compromised identity.
Measure: whether log coverage meets the investigation window you need, and the time to revoke all sessions and tokens for an identity.
Detect: would you find it, or would someone else tell you?
What it covers: continuous monitoring and analysis of adverse events.
Scenario: A customer’s security team tells you they’re seeing their data, which only your platform holds, on a criminal forum. Your own tooling has raised no alerts.
What to observe: how the report reaches the security team, how triage and severity are assigned, and which telemetry the team would need to reconstruct a slow exfiltration.
Measure: time to acknowledge, time to assign severity, and the source of detection (internal vs. external). Track the share of incidents found externally across exercises and real events.
Respond: can you contain it and communicate on time?
What it covers: incident management, incident analysis, reporting and communication, and mitigation.
Scenario: Ransomware is spreading from file servers to a hypervisor. Containing it means taking a revenue-critical system offline during business hours.
Injects to escalate:
- A reporter calls asking for comment.
- A key customer’s contract requires notification within a fixed window.
- Evidence suggests data was taken before encryption.
What to observe: who authorizes the containment trade-off, whether forensic evidence is preserved, and whether each notification clock is known and tracked. Examples: Form 8-K Item 1.05 for SEC registrants, due within four business days of a materiality determination; GDPR’s 72-hour notification to the supervisory authority; contractual commitments.
Measure: time to the containment decision, and time to each required notification compared with its deadline.
Recover: can you restore what matters, in the right order?
What it covers: carrying out the recovery plan and recovery communication.
Scenario: Restoration starts and the team finds the most recent backups were partly encrypted. The next clean restore point is older than your stated recovery point objective.
What to observe: whether restoration priorities were agreed in advance, who signs off that a restored system is clean, and how recovery status reaches customers and staff.
Measure: simulated time to restore tier-1 services against the recovery time objective (RTO), data loss against the recovery point objective (RPO), and whether the restoration order was documented beforehand.
If your organization uses AI tools or ships AI features, add AI incident scenarios to the Respond and Recover exercises. Most plans don’t account for them yet.
Summary: one scenario for each Function
| CSF 2.0 Function | Tabletop scenario | Primary measure |
|---|---|---|
| Govern | Holiday-weekend ransomware with the CISO unreachable | Time to formal incident declaration |
| Identify | Forgotten dev server holding production data | Time to identify owner and data classification; prior action items closed |
| Protect | MFA fatigue against a privileged admin | Log coverage vs. investigation window; time to revoke sessions |
| Detect | Customer reports your data on a forum | Time to acknowledge and assign severity; internal vs. external detection |
| Respond | Ransomware spreading to a revenue-critical system | Time to containment decision; notifications vs. deadlines |
| Recover | Partly encrypted backups | Restore time vs. RTO; data loss vs. RPO |
How do you evaluate incident response effectiveness?
You evaluate incident response effectiveness by setting measurable targets before you exercise or experience an incident, then comparing what actually happened against them. In incident response plan testing, use three kinds of measure: how fast decisions were made, how good the decisions were, and whether the findings were fixed.
- Set targets in advance. Write down the expected time to declare, escalate, contain, notify and restore before the exercise starts. Without a target, a timing is just a number.
- Measure decision speed. In a tabletop, technical work is simulated, so time the decisions rather than the technical steps: declaring the incident, escalating it, approving containment, and starting each notification.
- Measure decision quality. Record whether each decision was made by someone with documented authority, where the team departed from the plan, and what information the team needed but couldn’t get.
- Measure follow-through. Every after-action item needs an owner and a due date. Report the closure rate at the next exercise. A rising closure rate is the clearest sign the program is improving.
- Validate technical assumptions. A tabletop can’t prove your detections fire or your backups restore. Confirm those with technical testing, such as an assumed breach test or a restore test.
A simple way to score each Function after an exercise:
| Score | Meaning |
|---|---|
| 1: Not addressed | The plan doesn’t cover it, and participants improvised. |
| 2: Ad hoc | Participants knew what to do, but it isn’t written down. |
| 3: Documented | The plan covers it, and participants followed it. |
| 4: Practiced and measured | The plan covers it, the team met the target, and earlier findings are closed. |
What makes an incident response plan effective?
An effective incident response plan names who decides, defines what counts as an incident at each severity, and lists every notification obligation with its deadline. It’s also tested on a regular schedule and updated from what those tests find. In practice, that means the plan has:
- Named decision authority with deputies for declaring an incident, taking systems offline, engaging outside parties and approving disclosures.
- A severity model with examples so two people classify the same event the same way.
- A notification matrix covering regulators, customers, insurers, law enforcement and partners, with each deadline and the clock trigger that starts it.
- Out-of-band communication that still works if email, chat or identity systems are compromised.
- Scenario playbooks for your most likely incidents, such as ransomware, business email compromise, cloud account takeover and third-party compromise.
- Recovery priorities and objectives (RTO and RPO) agreed with the business before an incident.
- A link to risk management, as Rev. 3 recommends, so lessons learned change controls and priorities, not just the plan document.
- A testing cadence with tracked after-action items.
How often should incident response plan testing happen, and what do frameworks expect?
Run incident response plan testing at least once a year, and again after major changes: a new core system, an acquisition, a significant incident or a new regulatory obligation. Rotating through the six Functions over the year covers the whole framework without one marathon exercise.
Frameworks also expect evidence of testing:
- PCI DSS v4.0.1: Requirement 12.10.2 requires reviewing and testing the incident response plan at least once every 12 months.
- ISO/IEC 27001:2022: Annex A controls 5.24 to 5.28 cover incident management planning, assessing events, responding, learning from incidents and collecting evidence. Auditors usually ask for proof the plan has been exercised.
- SOC 2: The Common Criteria for system operations (CC7.3 to CC7.5) cover evaluating, responding to and recovering from incidents. Exercise reports and after-action tracking are common evidence.
[PLACEHOLDER: confirm the framework references against current versions with the compliance team before publishing.]
Where do CISA tabletop exercise packages fit?
CISA publishes free CISA Tabletop Exercise Packages (CTEPs) with ready-made scenarios, discussion questions and facilitator materials. They’re a good starting point for incident response plan testing, especially for a team running its first exercise. To use one for Rev. 3 testing, tag each module and discussion question with the CSF 2.0 Function it exercises. Then add injects specific to your environment, your regulators and your contracts, because generic packages can’t include those.
The main limitation is facilitation. When the person running the exercise also wrote the plan, gaps tend to get explained away instead of recorded. An independent facilitator keeps the findings honest.
How Canary Trap helps
Canary Trap runs tabletop exercises and incident response planning engagements for security and compliance teams, including incident response plan testing mapped to NIST SP 800-61 Rev. 3. We design scenarios around your environment, facilitate the session independently, and deliver a findings report with recommendations your team can act on.
Frequently asked questions
What is incident response plan testing?
Incident response plan testing checks whether your plan works when people actually have to use it. Usually that means a facilitated tabletop exercise, sometimes paired with technical tests such as restoring from backup or confirming a detection fires. The output is a set of measured gaps, each with an owner and a due date for the fix.
What are the top incident response frameworks?
The most widely referenced are NIST SP 800-61 Rev. 3 (organized around the six CSF 2.0 Functions), ISO/IEC 27035 (an international standard for information security incident management), and the SANS six-step model (Preparation, Identification, Containment, Eradication, Recovery, Lessons Learned). Many organizations base their plan on NIST and map it to ISO 27001 controls for certification.
What is the difference between NIST 800-61 Rev. 2 and Rev. 3?
Rev. 2 (2012) was a standalone incident handling guide built on a four-phase lifecycle. Rev. 3 (April 2025) is a CSF 2.0 Community Profile. It treats incident response as part of cybersecurity risk management and organizes it around Govern, Identify, Protect, Detect, Respond and Recover.
How fast should our incident response time be?
There’s no universal number. Set targets based on business impact and your obligations: regulatory notification windows, contractual commitments, and your recovery time and recovery point objectives. Then measure against those targets in every exercise and every real incident.
What’s involved in incident response training?
Training combines role-specific instruction (what each responder, executive and communicator does) with practice through tabletop exercises and technical drills. Practice matters most, because it shows where people’s understanding of the plan differs from what the plan says. See our tabletop exercise checklist for how to structure one.
What is the difference between incident response and disaster recovery?
Incident response handles a security event: detecting, analyzing, containing and recovering from it. Disaster recovery restores systems and data after any major disruption, security-related or not. In NIST Rev. 3, the Recover Function is where the two meet, so your recovery playbooks and your DR plan should agree on priorities.
How do you prioritize during incident response?
Prioritize by impact: how much the incident affects business functions, how sensitive the data involved is, and how recoverable the affected systems are. Those three factors come from NIST’s Rev. 2 guidance and still work as severity criteria. Define them with examples in your plan so the team doesn’t argue about severity mid-incident.