An emergency exercise is a structured rehearsal of the plan, the people, and the tools you claim will work in a disruption. The three types most programs mix up are not points on a vibe scale. A tabletop is discussion-only: roles, injects, and a hypothetical, with no resource deployment. A functional exercise runs a cell — an operations center, a department, a communications desk — on its real tools, while the field is often simulated. A full-scale exercise deploys people and kit into the space the plan is about. The design and after-action cycle used across U.S. public-sector exercises is HSEEP. The activity that validates whether a recovery plan works is a test plus an after-action that changes the plan.
Seminars and workshops sit beside this spectrum. They orient people to a plan. They do not validate it. A utility-shutdown procedure that must be revised is usually found in a functional or full-scale exercise, not in a seminar and not in a tabletop that never touches the procedure. This page is the type choice and the design cycle. How to write and facilitate the discussion itself is the tabletop design desk.
Pick the type by what has to move
| Type | What moves | Use |
|---|---|---|
| Tabletop | Talk and injects. No field deploy. | Policy, roles, first gaps |
| Functional | A cell and real tools; field often simulated | EOC, comms, a department |
| Full-scale | People and kit in the space | Prove the system when the risk earns it |
Tabletop: discussion, not deployment
What it means: Participants sit with their roles and work a hypothetical through injects — new information introduced on a clock. Nobody evacuates, nobody starts a generator, nobody moves a patient. The value is the argument: who decides, which plan they would open, what they do not know. An enhanced tabletop that adds a little simulation is still not a functional exercise and not a full-scale. If resources deploy, you have left the tabletop.
What good looks like: A scenario aimed at one decision the plan is fuzzy on, players who actually hold the roles, a facilitator who will not supply the answer, and notes that name the gap in the plan rather than praise the discussion. Time-boxed. The output is a list of plan changes, not a satisfaction score.
Common failure mode: The same attendees, the same scenario, and a write-up that says the plan worked because nobody disagreed out loud. Calling that write-up a full-scale, or an “enhanced” anything, does not change the fact that nothing moved.
Functional: the cell, the tools, a simulated field
What it means: A functional exercise activates one or more coordination cells and makes them use the plans, channels, and systems they would use for real. Controllers feed the field. Players do not. The emergency operations center tracks a simulated evacuation. The communications desk issues messages on the real notification path, to a controlled audience. A department runs the shutdown checklist against the real procedure, stopping short of the irreversible step if safety requires it — and the design says so in advance.
What good looks like: Objectives that name the cell and the tool. “Validate that the EOC can produce an accurate status within the first hour using the backup channel” is an objective. “Practice our response” is not. Evaluators watch the cell, not the slide deck. Failures of the tool — login, channel, stale contact list — are findings, even when the conversation was confident.
Common failure mode: A functional exercise run on email and a conference bridge that are not the bridges in the plan. The cell performs well on infrastructure the disruption would have removed. You have tested the meeting, not the function.
Full-scale: people and kit
What it means: A full-scale exercise deploys the people and the equipment the plan names, into the location the plan is about, under controllers and evaluators. Evacuation is walked. The generator is started or the transfer is attempted. Receiving sites, transport, and accountability are real movements, scaled to what is safe. This is the expensive type. Use it when the risk earns it: when a procedure can only be shown wrong by doing it, or when a regulator or a community partner expects a deployed exercise rather than a discussion.
What good looks like: A design that protects safety first, a scope that still stresses the objective, and partners who were in the planning and not surprised on the day. The after-action distinguishes what the deployment proved from what was still simulated. If a real emergency activates the plan, that activation can stand in for a deployed exercise only if you document it with the same honesty — what happened, what failed, what you changed.
Common failure mode: A full-scale that is a parade. Routes are pre-walked, equipment is staged by the control team, and the timeline is adjusted so the photo happens on schedule. Nothing in the plan is at risk of being wrong. The cost was real. The evidence was not.
Design it so it can fail: the HSEEP cycle
What it means: HSEEP — the Homeland Security Exercise and Evaluation Program — is FEMA’s cycle for building and judging exercises. In practice the cycle is: set objectives, build a scenario that can threaten those objectives, control the exercise so players are not coached, evaluate against the objectives, and write an after-action that becomes an improvement plan. The public program description lives with FEMA. You do not have to be a public agency to use the cycle. You do have to keep the steps in that order. A scenario written before the objectives will test whatever was dramatic, not whatever the plan needs.
What good looks like: A short design meeting that produces objectives a stranger could score yes or no. Scenario injects exist only to force those objectives. Controllers know what they may tell players and what they must not. Evaluators have the plan excerpt in front of them. The improvement plan names an owner and a date for each change, and the next exercise checks whether the change was made.
Common failure mode: Objectives written after the exercise so that they match what happened. The after-action is a narrative. There is no improvement plan, or the improvement plan is a list with no owner. Next year the same gap is facilitated again and recorded as engagement.
The after-action is the product
What it means: An after-action report records what the exercise was trying to show, what players did, where that diverged from the plan, and what will change. The improvement plan is the list of those changes with owners. Without the change, the exercise was a meeting. This is true of tabletops as much as full-scale deployments. Discussion that surfaces a broken call tree has done its job only when the call tree is replaced.
What good looks like: Findings tied to an objective, evidence a player action or a tool failure, and a plan edit you can point to afterward. Hot wash the same day, while the disagreement is specific. Write the report before the memory becomes a compliment. Track the improvement items in the same place you track plan revisions, so an auditor or a surveyor can see that the exercise moved the document.
Common failure mode: A sign-in sheet and a paragraph that says the exercise was a success. No inject log, no evaluator notes, no revised page in the plan. For facilities that must show training and testing, that file reads as theater.
Who needs the harder type
What it means: Match the type to the claim you need to make. If the claim is “the roles are clear,” a tabletop can carry it. If the claim is “this cell can operate on its backup channel,” you need a functional. If the claim is “we can move these people, with this equipment, to that receiving site,” you need a deployment. Healthcare providers under the CMS emergency preparedness program have a testing cadence that is more specific than a generic annual tabletop — the CMS four-components guide lays out the exercise expectations and the after-action loop surveyors read. Do not assume one cadence fits every provider type.
What good looks like: A yearly mix that includes at least one exercise capable of finding a procedural error, not only a policy argument. Communication is exercised on the channels the crisis plan names, including one path that is not the everyday mail system. Partners who appear in the plan — a receiving facility, a transporter, a local emergency manager — are in an exercise often enough that their phone number is real.
Common failure mode: Two tabletops a year with the crisis team, counted as the whole program. The night shift, the receiving site, and the actual notification system are never in the room. The plan’s hardest page has not been looked at since it was pasted in.
FAQ
What is a tabletop exercise?
A discussion-based review of roles on a hypothetical, driven by injects. It is not a full-scale deployment and not a functional exercise under another name. No resources are deployed to the field.
Which type of exercise is discussion-based?
The tabletop. Seminars and workshops are also discussion, and they are lighter still: they orient people to content. They are not validation. Functional and full-scale exercises are operations-based.
What is the difference between a functional and a full-scale exercise?
A functional exercise runs a cell and its real tools while the world outside that cell is simulated. A full-scale exercise puts people and equipment into play. If you did not deploy people and kit, you did not run a full-scale, however large the conference room was.
What is HSEEP?
FEMA’s exercise design and evaluation cycle: objectives, scenario, control, evaluate, improvement plan. The point of the cycle is an after-action that changes the plan. A design that cannot fail the objectives is not finished. FEMA publishes the program at fema.gov.
Where do we find out a procedure is wrong?
Usually in a functional or full-scale exercise, when someone has to execute the procedure rather than describe it. A tabletop can reveal that nobody agrees who owns the step. It rarely reveals that the step itself cannot be done — the valve, the bus, the contact list, the oxygen, the badge access. That shows up when something moves.
How this connects
This page sits under the emergency preparedness guide and beside the tabletop design desk, which is where facilitation technique lives. Message content, incident command, and who is allowed to speak are in the crisis communication protocols guide. For healthcare, the regulatory frame is the CMS Emergency Preparedness Rule and the broader healthcare continuity compliance guide covering CMS, Joint Commission, and HIPAA contingency planning. An exercise that never touches the communication plan or the healthcare testing cadence has not tested the program those pages describe.