Schedule lags are appropriate when a relationship requires a real, measurable passage of time and no work, resources, or meaningful progress occurs during that interval. A fixed curing, drying, cooling, or stabilization period may justify a lag. However, if the interval includes work, ownership, status, cost, uncertainty, or a deliverable, it should usually appear as an activity or milestone.
Lags should remain the exception rather than the standard way to position activities. They can hide schedule detail, complicate status, affect float, and weaken schedule risk analysis. Therefore, use a lag only when it represents the technical nature of the dependency—not because it produces a preferred date.
What Is a Schedule Lag?
A lag is a positive time offset applied to a logical relationship between a predecessor and successor. For example, a finish-to-start relationship with a three-day lag prevents the successor from starting until three days after the predecessor finishes.
The lag belongs to the relationship, not to either activity. It does not create a separately named task with its own owner, resources, start, finish, or status. Microsoft describes lag time as a delay between dependent tasks and allows users to enter a positive value in the relationship’s lag field. Its guidance for adding lead or lag time in Microsoft Project shows an entry such as 2d for a two-day lag.
A lead uses a negative offset to create overlap. Leads raise different concerns because they can model work beginning before its predecessor finishes. See why leads can damage schedule quality for a focused discussion of that issue.
When Schedule Lags Are Appropriate
A defensible lag normally meets all of the following conditions.
The interval represents passage of time rather than work
The strongest use case is a physical process that must run for a defined period after one activity and before another. Examples include:
- Concrete curing before forms can be removed
- Coating or adhesive curing before inspection
- Thermal stabilization before calibration
- A fixed equipment soak period before testing
- Cooling or drying time before handling
NASA identifies examples such as concrete cure time and printed circuit board coating bake-out time in its Schedule Management Handbook. However, the handbook also recommends replacing most lags with clearly labeled activities because activities provide better visibility and status capability.
The delay has a sound technical basis
A lag should have a documented basis such as an engineering procedure, test requirement, manufacturer instruction, or approved planning assumption. Avoid arbitrary entries such as FS+10d when the scheduler cannot explain why ten days must pass.
The duration should also remain reasonably stable. If the interval could range from three to fifteen days, a fixed lag may create false precision. Instead, consider a separate activity that can carry duration uncertainty and receive status.
There is nothing useful to status during the interval
A pure waiting period does not produce intermediate progress. Consequently, the team gains little by reporting it as 30, 60, or 90 percent complete.
However, many apparent waiting periods include real work. A customer review may involve comment development, adjudication, revision, and approval. Shipping may require handling, transit tracking, receipt, and inspection. Procurement lead time may include fabrication, supplier testing, and delivery. In those cases, separate activities usually provide a more accurate plan.
The lag uses the correct calendar
Calendar selection can materially change the result. A five-day working-time lag does not produce the same date as five elapsed calendar days when weekends or holidays intervene.
Deltek Open Plan allows a calendar to be assigned to a relationship. Its relationship guidance explains that Open Plan otherwise uses the successor activity’s calendar and gives a 24-hour calendar as an example for concrete curing. Schedulers should verify the lag calendar and time unit before baselining or delivering the schedule.
When to Replace a Lag With an Activity
The U.S. Government Accountability Office advises schedulers to use lags only for the passage of time, make every effort to avoid them through better activity decomposition, and document compelling reasons for those that remain. These principles appear in the GAO Schedule Assessment Guide.
Replace a lag with an activity when any of the following applies:
- Someone performs work during the interval.
- The interval consumes labor, material, facilities, or other resources.
- A control account manager needs to report progress.
- The duration may change based on performance or risk.
- The interval produces an identifiable output or approval.
- Management needs visibility into the responsible organization.
- The schedule risk analysis should apply an uncertainty distribution.
- The customer requires the event to appear explicitly in the Integrated Master Schedule.
For example, do not hide a ten-day engineering review behind an FS+10d relationship. Create an activity such as Perform System Requirements Review, followed by an approval milestone if approval represents a distinct event. The result gives the team something to assign, status, analyze, and forecast.
Why Lags Matter to Critical Path and Float
Lags participate in the schedule network calculation. Therefore, a lag on a driving relationship can delay the successor, consume float, or become part of the longest path to a milestone.
However, the relationship offset may not be obvious in a standard activity table or bar chart. A reviewer can see an unexplained gap between activities without immediately knowing whether a lag, constraint, calendar, or missing activity caused it. That reduces transparency during critical path analysis.
Large or numerous lags can also make it harder to identify the driving predecessor. In addition, they may create confusing float results when combined with constraints and multiple calendars. For background on these calculations, see free float versus total float and the Critical Path Method guide for program schedulers.
A lag should never serve as a substitute for contingency. Adding FS+20d because an activity might run late does not model the risk event, probability, or consequence. Instead, it silently shifts downstream dates. Address uncertainty through risk-informed durations, explicit risk mitigation work, schedule risk analysis, or an approved schedule reserve approach.
Schedule Lags in EVMS and Schedule Risk Analysis
A lag may move the planned dates of budgeted successor work, but the lag itself does not represent scope. As a result, it has no separately identifiable budget, actual cost, or earned value.
This distinction matters in an Earned Value Management System (EVMS). A long lag can place a substantial gap in a control account plan without showing what management should monitor during that period. It can also obscure whether a forecast delay resulted from technical work, external dependency, risk, or a planning assumption.
NASA warns that lags cannot be statused like normal activities and that uncertainty distributions cannot be applied to them for schedule risk analysis. Consequently, an uncertain wait modeled as a fixed lag may understate forecast uncertainty. If the duration could vary materially, model it as an activity and assign a supportable uncertainty range during the Schedule Risk Analysis (SRA).
A Practical Decision Test for Schedulers
Before entering a lag, ask these five questions:
- Does the interval represent only the passage of time? If work occurs, use an activity.
- Is the duration fixed and technically supported? If it varies with performance, use an activity.
- Is there truly nothing to resource or status? If someone owns progress, use an activity.
- Can the scheduling tool calculate the correct calendar basis? Confirm working time versus elapsed time.
- Can the team document and defend the lag? Record the basis in task notes, a custom field, or the schedule basis document.
If the answer to any of the first four questions is no, an activity will usually produce a stronger schedule. In addition, periodically filter the schedule for non-zero lag values. Revalidate each one during baseline development, monthly status, replanning, and customer delivery.
Fictional Program Example: Coating Cure Time
Assume the fictional Northstar Relay Demonstration includes an electronics assembly that receives a conformal coating. The approved process requires 24 elapsed hours of cure time before quality inspection can begin.
The scheduler could create this relationship:
- Apply Conformal Coating
- Finish-to-start relationship with a 24-hour elapsed lag
- Inspect Coated Assembly
This lag may be defensible if the cure period requires no labor, produces no intermediate progress, and remains fixed under the applicable procedure. The scheduler should document the procedure reference and verify that the relationship uses an elapsed-time calendar.
Now assume technicians must monitor temperature and humidity, record readings every four hours, correct excursions, and obtain a quality release. The interval now includes accountable work and a completion event. A better model would include Monitor Coating Cure Cycle followed by a Quality Release milestone. That structure improves ownership, status, risk analysis, and forecast credibility.
Using Lags in Microsoft Project and Deltek Open Plan
Microsoft Project
Microsoft Project accepts positive lag in the Lag column or as part of the predecessor entry. For example, 42FS+3d creates a finish-to-start relationship with a three-day delay. Microsoft also permits percentage-based lags.
Use percentage lags cautiously. If the predecessor duration changes, the offset can change as well. A fixed technical wait is usually easier to understand and audit when entered in a defined time unit. Also confirm calendars and inspect the calculated dates rather than assuming that 3d represents three consecutive calendar days.
Deltek Open Plan
Open Plan supports positive and negative relationship lags as well as relationship calendars. Because its forward and backward passes consider relationship lags, those values can affect early dates, late dates, float, and criticality.
Use a relationship calendar when the waiting period follows a different calendar from the successor. For example, a physical cure process may continue around the clock even though the inspection team works a standard weekday calendar.
Proposal and Customer-Delivery Considerations
Proposal schedules often contain lags because teams build them quickly and want to avoid adding detail. However, an unexplained lag can look like hidden work, schedule padding, or date manipulation during source selection.
The Department of Defense Integrated Master Plan and Integrated Master Schedule Preparation and Use Guide, available through Defense Acquisition University, advises reviewers to determine whether lags are reasonable and realistic or are being used to drive a date. It also recommends explaining assumptions when lags represent anticipated delays.
Before delivery, include lag review in the broader IMS customer-delivery review. At minimum, verify that each lag has a technical basis, correct calendar, current duration, and documented justification.
Best Practice Is Not Automatically a Contract Requirement
GAO, NASA, and DoD scheduling guides strongly favor limited, justified lag use. However, guidance does not automatically create a universal contractual prohibition against lags.
The controlling requirements come from the applicable solicitation, contract, statement of work, Contract Data Requirements List, data item description, scheduling specification, and approved program procedures. Some customers may impose thresholds, require narrative justification, or prohibit lag without approval. Therefore, review the governing documents before establishing proposal or execution-schedule rules.
Frequently Asked Questions
Are lags prohibited in a DoD Integrated Master Schedule?
Not as a universal scheduling rule. However, authoritative guidance recommends minimizing and justifying them, while a specific solicitation or contract may impose stricter requirements.
Does a lag affect the critical path?
Yes. Scheduling engines include relationship lags in network calculations. A lag on a driving path can delay downstream activities and affect total float.
Should concrete cure time be a lag or an activity?
Either may be supportable. A fixed, unattended cure period can justify a lag. If the period requires monitoring, testing, resources, variable duration, or formal release, use one or more activities.
Is a lag the same as schedule contingency?
No. A lag models a defined delay between two logically related events. Contingency or schedule reserve addresses uncertainty and should remain visible, controlled, and consistent with program procedures.
Final Interpretation
Use schedule lags only for genuine, supportable waiting time that contains no work and needs no meaningful status. Keep the duration specific, select the correct calendar, document the basis, and confirm that the lag does not conceal risk or scope.
When in doubt, create an activity. Explicit activities make the schedule easier to execute, status, analyze, explain, and defend during an Integrated Baseline Review, schedule risk analysis, proposal evaluation, or customer delivery.