Introduction
You report the retention rate in Part E of the Fall Enrollment (EF) component, which IPEDS collects in its spring collection window, and it ends up on College Navigator next to your institution's name.
The formula is one line. The work is in the cohort: which students belong in it, the four reasons you may remove someone, the single reason you may add someone back, and the date you count on. Everything below comes from the 2025-26 Fall Enrollment survey materials and the IPEDS glossary. Check the current year's instructions in the IPEDS Data Collection System before you submit, since NCES revises wording each cycle.
If you are newer to the federal surveys, IPEDS in plain English covers the components and how they fit together.
The formula
For the 2025-26 collection, which measures the Fall 2024 cohort a year later, NCES gives the calculation as:
Retention rate = (students from the Fall 2024 cohort still enrolled in Fall 2025 + students from the Fall 2024 cohort who completed their program by Fall 2025) ÷ adjusted Fall 2024 cohort × 100
The adjusted cohort is the original cohort minus allowable exclusions plus allowable inclusions. You calculate it twice, once for the full-time cohort and once for the part-time cohort, and the IPEDS system computes the percentage when you save each screen. Shift every year forward by one for the next collection.
Who is in the cohort
- First-time students only. The IPEDS glossary defines a first-time undergraduate as a student with no prior postsecondary experience attending any institution for the first time at the undergraduate level. Transfer-ins are not in the retention cohort.
- Degree or certificate-seeking only. Non-degree students are out.
- At four-year institutions, bachelor's-seeking students only. The cohort is the first-time bachelor's (or equivalent) degree-seeking group, so at an institution that also awards associate's degrees or certificates, it is a subset of the first-time total and should be less than or equal to it.
- At two-year and less-than-two-year institutions, all first-time degree or certificate-seeking students.
- Full-time and part-time are separate cohorts, fixed by attendance status in the first fall. A student who was full-time in Fall 2024 stays in the full-time cohort even if they are part-time in Fall 2025. For undergraduates, full-time means 12 or more semester or quarter credits, or 24 or more clock hours a week, each term.
- The cohort is preloaded. IPEDS carries the full-time and part-time first-time counts forward from Part A of your prior year Fall Enrollment submission. You verify it; you do not rebuild it. If the preloaded number looks wrong, the error is in last year's Part A.
Exclusions and the one inclusion
- Exclusion 1: the student died or became totally and permanently disabled.
- Exclusion 2: the student left to serve in the armed forces, including being called to active duty.
- Exclusion 3: the student left to serve with a foreign aid service of the federal government, such as the Peace Corps.
- Exclusion 4: the student left to serve on an official church mission.
- The one inclusion: if your institution left freshman study abroad students out of last year's first-time count, add back those who re-enrolled at your institution in their second year.
That is the whole list. Students who transferred to another college, stopped out for work, or took a leave of absence are not exclusions. They stay in the denominator and count as not retained.
A related study abroad rule runs the other way: sophomores studying abroad this fall may be missing from your fall enrollment count, but they still count as retained.
Who counts as retained
- Students from the cohort enrolled for credit at your institution in the second fall. Enrollment at another institution does not count.
- Students from the cohort who completed their program by the second fall. At a four-year institution this means completed the bachelor's degree, which NCES expects to be zero or very small.
- Count a student once. If a two-year student finished a certificate and is now enrolled in another program, they are retained one time, not twice.
- Leave of absence is not retained. NCES is explicit: a student must be enrolled for credit in the fall to count.
- Use the right date. Academic reporters count as of the institution's official fall reporting date or October 15. Program reporters build the cohort from students enrolled any time between August 1 and October 31 of the first year and count retention as of August 1 of the next year.
Worked example: a four-year institution
| Line | Full-time cohort | Part-time cohort |
|---|---|---|
| First-time bachelor's-seeking cohort, Fall 2024 (preloaded) | 1,200 | 150 |
| Exclusions (2 armed forces, 2 church missions, 1 died) | 5 | 0 |
| Inclusions (freshman study abroad students who re-enrolled) | 4 | 0 |
| Adjusted cohort (1,200 minus 5 plus 4) | 1,199 | 150 |
| Enrolled for credit on census date, Fall 2025 | 902 | 72 |
| Sophomores studying abroad, Fall 2025 | 10 | 0 |
| Completed a bachelor's degree by Fall 2025 | 0 | 0 |
| Total retained | 912 | 72 |
| Retention rate | 76.1% | 48.0% |
These numbers are illustrative, not from any real institution. The full-time rate is 912 ÷ 1,199 × 100 = 76.1%. The part-time rate is 72 ÷ 150 × 100 = 48.0%.
Two details carry most of the risk. First, 287 full-time students were not retained. If some of them transferred to a nearby university after one year, they still stay in the denominator; transfer is not an exclusion. Second, the 10 sophomores abroad are counted as retained even though your Fall 2025 headcount file may not list them, so the retained count does not come from a simple join to this fall's census.
Worked example: a two-year institution
Again with illustrative numbers. A community college's full-time, first-time degree or certificate-seeking cohort from Fall 2024 is 800 students. Two left for active military duty, so the adjusted cohort is 798. On the Fall 2025 census date, 410 cohort members are enrolled for credit. Another 38 completed a one-year certificate during 2024-25 and did not return; they count as retained because they completed their program.
Retention rate = (410 + 38) ÷ 798 × 100 = 56.1%.
If 12 of the 410 enrolled students also completed a certificate along the way, they are already in the 410. Adding them again would overstate the rate. This is the double-count NCES warns about for less-than-four-year institutions.
Common mistakes
Treating transfers-out as exclusions
The exclusion list has four reasons and transfer is not one of them. Students who leave for another college count against your retention rate. Transfer-out shows up in the Graduation Rates component, which is a different survey with a different cohort.
Moving students between cohorts
Full-time or part-time status is fixed by the first fall. Re-sorting students by their second-fall status changes both rates and does not match the preloaded cohort.
Using the whole first-time group at a four-year institution
Four-year institutions report the bachelor's-seeking cohort. Associate's and certificate seekers are not in it.
Counting on the wrong date
An end-of-term enrollment snapshot includes late adds and excludes early withdrawals that the census date would not. Use the official fall reporting date or October 15, consistently with the cohort year.
Publishing an internal number under the IPEDS name
Many offices also track term-to-term or all-new-student retention. Those are useful, but label them, because a reader comparing your number to College Navigator will assume the IPEDS definition.
Retention is not persistence
IPEDS retention only credits enrollment at your own institution. The National Student Clearinghouse Research Center reports both measures and draws the line clearly: in its persistence and retention report, students are retained if they remain at their starting institution and persisted if they remain enrolled at any institution. A student who transfers after one year is a persistence success and a retention loss.
That gap is why the same campus can quote three different "retention" figures in one cabinet meeting. Our scenario post on what happens when retention means five different things walks through settling it, and predicting term-to-term retention covers the early-warning version of the metric.
If you want to check your rate against peers, the published IPEDS retention rates for every institution are in the federal data. With Clema for IPEDS you can ask for the full-time retention rate of your peer group in plain English and get the number with its collection year attached.
Sources
- NCES, 2025-26 IPEDS Fall Enrollment survey materials: Part E screens, instructions, and FAQs (formula, exclusions, inclusions, study abroad, leave of absence, reporting dates).
- NCES, IPEDS glossary: retention rate, first-time student, full-time student, adjusted cohort.
- Association for Institutional Research, Fall Enrollment retention rates tutorial for four-year institutions (2024-25 collection).
- NCES, IPEDS survey components and data collection cycle.
- National Student Clearinghouse Research Center, Persistence and retention.
Compare your retention rate to peers in one question
Clema answers IPEDS questions in plain English and shows the survey component and collection year behind every number.
Book a demo