The DCMA high duration check identifies unfinished schedule activities with a baseline duration greater than 44 working days. Under the commonly used Defense Contract Management Agency (DCMA) 14-point schedule assessment, high-duration activities should generally represent less than 5 percent of the applicable incomplete activity population.
A result above the threshold does not automatically make the schedule unacceptable. Instead, it signals that the scheduler and Control Account Manager (CAM) should determine whether the activity hides measurable handoffs, weakens status visibility, or makes progress too subjective. Some long activities are valid, but they need a defensible basis.
What the DCMA high duration metric measures
The high-duration check focuses on the length of unfinished activities. The Department of Defense Risk, Issue, and Opportunity Management Guide describes high-duration activities as unfinished tasks with a baseline duration greater than 44 working days. It presents a goal of less than 5 percent as part of a sample 14-point schedule health assessment.
The current DAU Project and Program Management Fundamentals Handbook adds useful detail. It identifies incomplete tasks with a baseline duration greater than 44 working days and a baseline start within the detail planning period or rolling wave. Therefore, a program should confirm the exact assessment population defined by its customer, schedule management plan, system description, or assessment tool.
The basic calculation is:
High-duration percentage = incomplete qualifying activities over 44 working days ÷ total incomplete qualifying activities × 100
For example, a schedule with 180 incomplete qualifying activities and eight high-duration activities has a rate of 4.44 percent. It falls below the commonly used 5 percent threshold. If 12 activities exceed 44 working days, the rate becomes 6.67 percent and warrants further review.
The 44-day threshold is an assessment criterion, not a universal contract clause
The 44-working-day threshold is a schedule health metric and planning convention. It is not, by itself, a universal Federal Acquisition Regulation or Defense Federal Acquisition Regulation Supplement requirement.
For applicable DoD contracts, DFARS Subpart 234.2 addresses when an Earned Value Management System (EVMS) must comply with the EIA-748 guidelines. The associated EVMS clause also requires management procedures that generate timely, reliable, and verifiable information for required performance reports and the Integrated Master Schedule (IMS). However, those provisions do not establish 44 working days as a universal maximum activity duration.
A contract, Contract Data Requirements List, Integrated Program Management Data and Analysis Report requirement, schedule preparation instruction, or negotiated program procedure may establish more specific rules. An agency or customer may also tailor the threshold. Therefore, the contract and incorporated requirements control.
Programs should treat the DCMA check as a diagnostic unless the threshold has been made contractually binding through specific contract language. For broader context, see the complete DCMA 14-point schedule assessment guide.
Why long activities create management problems
Long activities often reduce visibility into the work. A task named “Complete subsystem design” with a duration of 90 working days may include requirements analysis, preliminary drawings, peer reviews, analysis updates, drawing release, and approval. If the schedule models all that work as one activity, management cannot see the internal handoffs or determine which step caused a delay.
The GAO Schedule Assessment Guide recommends reasonably short, meaningful durations that support discrete progress measurement. It also explains that planners should divide long activities when they can identify logical breaks. However, GAO cautions against excessive detail because very short activities can create an unnecessary maintenance burden.
High-duration activities can cause several practical problems:
- Subjective status: The CAM may report 40 or 70 percent complete without an objective completion event supporting that value.
- Late problem detection: A missed internal handoff may remain hidden until the activity approaches its forecast finish.
- Weak logic: Successor activities may depend only on the final completion of a broad work package even though some downstream work could begin earlier.
- Poor forecast quality: The schedule cannot show which remaining step drives the activity finish.
- Limited accountability: A single activity may span several responsible teams, facilities, or suppliers.
- Weak schedule risk analysis: A broad duration can combine several different uncertainty drivers into one estimate.
Long activities do not necessarily distort the critical path mathematically. However, they can hide the work and assumptions inside the path. As a result, management may understand the forecast finish date without understanding what drives it.
Baseline duration, current duration, and remaining duration are different
The standard high-duration metric generally uses baseline duration. It does not simply test the current or remaining duration shown during the latest update.
Suppose an activity had a baseline duration of 65 working days. It has now been in progress for 50 days and has 15 days remaining. The activity still meets a baseline-duration definition of high duration even though its remaining duration is below 44 days.
Conversely, an activity may have a baseline duration of 30 days but grow to a current duration of 60 days after status updates. It may not fail a strictly baseline-based calculation. However, the growth is still a significant forecasting issue. The scheduler should investigate the variance, remaining work, logic, and estimated completion date.
Programs often supplement the formal metric with current-duration or remaining-duration analysis. That approach can reveal execution problems, but it should not be mislabeled as the baseline-duration check. The assessment report should state which field and population it tested.
Which schedule records should be included?
The calculation should focus on incomplete detail activities that represent measurable work. Summary rows and zero-duration milestones do not belong in a duration test. Level of effort, planning packages, and other non-discrete records may also require separate treatment.
The GAO guide advises analysts to exclude nonapplicable records such as level-of-effort and reference activities when assessing detail activity duration. Planning packages may legitimately span longer periods until the program converts them into detailed work packages and activities. Therefore, an unfiltered test of every row can produce a misleading result.
Before calculating the metric, document how the analysis handles:
- Completed activities
- Summary activities
- Milestones
- Level-of-effort activities
- Planning packages outside the detail planning window
- External or reference activities
- Inactive activities
- Activities assigned to nonstandard calendars
Consistency matters. The monthly assessment should use the same documented population unless the program formally changes its method.
When a high-duration activity may be valid
Schedulers should not divide work only to improve a metric. Some activities have valid long durations because the work lacks meaningful intermediate handoffs or measurable completion points.
Examples may include environmental exposure testing, material aging, supplier fabrication, extended qualification testing, regulatory review, or an external approval cycle. Even then, the team should determine whether the activity includes separately identifiable events such as test setup, test execution, data analysis, report preparation, approval, shipment, or receipt.
A valid exception should have a clear basis. The activity description should identify the actual work or process rather than use a broad label. In addition, the schedule basis document or narrative should explain why further decomposition would not improve planning, status, or control.
Long level-of-effort activities require different treatment. Program management, configuration management, and recurring engineering support may span many months because their duration follows other discrete work. They should not be split into artificial monthly tasks merely to satisfy a discrete-work metric. However, they also should not drive the critical path as if they represented product completion.
How to correct a high-duration activity
First, ask the responsible CAM or technical lead what observable events occur between the activity start and finish. The best decomposition points usually come from the work itself, not from an arbitrary duration target.
Useful breakpoints include:
- Completion of a technical product
- Transfer of work between responsible organizations
- Submission, review, and approval events
- Supplier release, fabrication, inspection, and delivery
- Entry and exit criteria for a test phase
- Availability of an input needed by another team
- A point where objective progress can be measured
Next, connect the new activities with valid logic. Do not automatically create a simple finish-to-start chain if the work overlaps or has different technical dependencies. Review the guidance on DCMA relationship types and finish-to-start logic before restructuring the network.
Finally, confirm that each new activity has a clear scope, responsible owner, realistic duration, appropriate calendar, and objective completion condition. Check for missing predecessors and successors after making the change. Artificial decomposition can create new open ends, so rerun the DCMA missing logic check.
Fictional example: decomposing a 75-day design activity
A fictional avionics development program includes the activity “Develop Interface Control Document,” with a baseline duration of 75 working days. The activity starts after interface requirements become available and finishes before software integration begins.
During review, the scheduler learns that the work includes four distinct outcomes:
- Develop the initial interface definition.
- Conduct an internal engineering review.
- Resolve comments and release the document.
- Obtain customer concurrence.
The team replaces the 75-day activity with four activities and a concurrence milestone. It also links the initial approved sections to software planning work that does not need to wait for final customer concurrence.
This change improves more than the metric. The program can now identify the current design stage, assign ownership to each handoff, measure completed work objectively, and evaluate the effect of late comments. In addition, the network shows which approval actually drives software integration.
However, the scheduler should not divide the work into three arbitrary 25-day tasks named “Part 1,” “Part 2,” and “Part 3.” Those labels do not describe measurable products or meaningful dependencies.
Checking high duration in Microsoft Project
Microsoft Project defines duration as the span of active working time between a task’s scheduled start and finish. It calculates that span using task dependencies, constraints, and applicable calendars. Microsoft explains this behavior in its documentation for the Duration task field.
For a baseline-based high-duration review in Microsoft Project, display at least the task name, outline level, active status, percent complete, baseline start, baseline finish, baseline duration, current duration, remaining duration, task calendar, and task type. Then filter the records according to the program’s documented assessment population.
Do not count calendar days between the baseline start and finish. The metric refers to working duration. Also confirm the project’s hours-per-day settings and calendars because Project converts displayed duration units according to its scheduling configuration.
After correcting activities, recalculate the schedule and inspect the logic, critical path, total float, and milestone dates. Then perform a normal status-quality review using the practices in How to Status an Integrated Master Schedule.
Common mistakes during the review
- Treating 44 days as an absolute prohibition: The threshold prompts analysis. It does not prove every longer activity is invalid.
- Using current duration instead of baseline duration: That changes the metric unless the customer has defined a different test.
- Counting summary rows and milestones: These records distort the denominator and do not represent comparable detail work.
- Splitting activities into arbitrary pieces: Decomposition should expose products, handoffs, or measurable events.
- Ignoring the detail planning window: Long planning packages outside the rolling wave may require separate treatment.
- Passing the metric without reviewing exceptions: A schedule can remain below 5 percent while still containing a critical 200-day activity with weak status visibility.
- Changing the baseline without control: Correcting future schedule detail does not justify rewriting approved baseline history outside the program’s change-control process.
Practical interpretation for schedule reviews
The percentage provides a useful screening result, but the exception list provides the real management value. Review the longest activities first, especially those on critical or near-critical paths, those with significant budget, and those tied to high-risk technical work.
For each exception, determine whether the activity has a clear outcome, objective status method, valid duration basis, responsible owner, and appropriate logic. Also ask whether another team needs an intermediate product before the activity finishes.
A mature review does not aim merely to turn the dashboard green. It uses the DCMA high duration check to improve forecast credibility and execution visibility while preserving valid long-duration work where decomposition would add no value.
Frequently asked questions
Does a 44-working-day activity fail the check?
No. The commonly cited definition identifies activities with a baseline duration greater than 44 working days. A task at exactly 44 working days does not exceed that threshold.
Should every activity be shorter than the reporting period?
Not necessarily. Shorter activities often improve monthly status visibility, but activity length should follow the nature of the work. The schedule should balance measurable detail against the effort required to maintain it.
Can a program use a threshold other than 44 days?
Yes. A customer, agency, contract, or program procedure may define another value. The schedule management plan should document the threshold and calculation method.
Does passing the high-duration check prove the schedule is healthy?
No. It addresses only one schedule characteristic. Logic, constraints, float, status quality, critical path validity, resource assumptions, and baseline execution also affect schedule reliability.