Consider the rower first. Nine of the twenty-six elite Division III colleges in this analysis field rowing at all. At those nine, rowing accounts for about 20 roster places per entering class, one of the three largest per-college squads in the division. Seventeen of the twenty-six colleges offer nothing.
Now the golfer. Twenty-one of the twenty-six colleges sponsor golf, so almost every school on the list is live. But golf accounts for about 4.4 roster places per entering class at each of them. A coach with four positions has a short list and a long memory.
Same student profile, same target colleges, opposite problems. The rower has depth and no breadth. The golfer has breadth and no depth. Neither family learns this from a college website, because no college publishes roster capacity in a form that can be compared against another college.
This is the second article in a series on athletics and admissions at selective Division III colleges. The first mapped how much roster capacity each college carries. This one takes the same 26 colleges and asks which sports that capacity actually sits in, using the sport-level Equity in Athletics Disclosure Act file for 2024-25.
Two cautions carry through the whole piece. The first is the one that governed the previous article: these are participation counts, not recruiting offers. A roster place counts everyone on the team, including walk-ons and students who joined after arriving on campus.
The second is specific to this data and it changes how two sports read. Athletes who compete in more than one season are counted more than once in the federal sport file, which inflates track and cross country and, to a lesser degree, swimming and diving. Comparing the sport file against unduplicated headcounts shows 20.5 percent more roster entries than there are distinct athletes across these 26 colleges. Both figures are reported below, and the lower one is used in every comparison.
Track and Cross Country Is the Largest Sport, but Not by as Much as the Raw Data Says
Adding every roster across all 26 colleges and dividing by four class years gives the number of places each sport accounts for in a single entering class.
Estimated varsity roster places per entering class by sport, summed across 26 elite Division III colleges. Data: EADA sport-level file, 2024-25. Track and cross country and swimming and diving are normalized into single categories because colleges report them inconsistently.
Read the sport file at face value and track and cross country accounts for roughly 1,396 roster places per entering class against football’s 435. That comparison is not sound. Distance runners at these colleges appear on cross country, indoor track and outdoor track rosters, and the federal file counts them separately each time. Football has one season and no equivalent problem.
Allocating the duplication to track and to swimming puts track at about 789 places per entering class, still the largest sport in elite Division III but by a factor closer to two than three. Soccer at 394 and lacrosse at 368 move up the ranking, and swimming and diving falls from 387 to about 222 for the same reason track does. Those two adjusted figures are estimates at the low end of a range, because some of the duplication belongs to students who play two different sports rather than two seasons of one.
Football is worth understanding precisely because of what it is not. Twenty of the 26 colleges field it, and at 21.7 places per entering class it is one of the largest single squads where it exists. But it is a men’s sport with no women’s counterpart, six of these colleges do not field it, and even on the adjusted comparison it sits at roughly half of track.
The full table gives the working numbers.
| Sport | Colleges sponsoring | Places per class | Places per class, per college |
| Track and cross country | 26 of 26 | 789 est. (1,396 unadjusted) | 30.4 |
| Football | 20 of 26 | 435 | 21.7 |
| Soccer | 26 of 26 | 394 | 15.2 |
| Lacrosse | 19 of 26 | 368 | 19.4 |
| Baseball | 25 of 26 | 223 | 8.9 |
| Swimming and diving | 25 of 26 | 222 est. (387 unadjusted) | 8.9 |
| Basketball | 26 of 26 | 201 | 7.7 |
| Rowing | 9 of 26 | 177 | 19.6 |
| Tennis | 26 of 26 | 152 | 5.8 |
| Volleyball | 26 of 26 | 128 | 4.9 |
| Ice hockey | 9 of 26 | 119 | 13.2 |
| Field hockey | 18 of 26 | 109 | 6.0 |
| Softball | 23 of 26 | 107 | 4.7 |
| Squash | 14 of 26 | 94 | 6.7 |
| Golf | 21 of 26 | 93 | 4.4 |
| Wrestling | 7 of 26 | 51 | 7.3 |
| Fencing | 5 of 26 | 45 | 9.0 |
| Skiing | 5 of 26 | 34 | 6.8 |
Sponsorship and roster places by sport across 26 elite Division III colleges, 2024-25. Track and swimming figures are estimates that allocate all cross-roster duplication to those two categories, with the unadjusted count shown in parentheses. Sports sponsored by fewer than three colleges are omitted. Data: EADA sport-level file and institution-level unduplicated counts, 2024-25.
Wide Doors and Deep Doors Are Rarely the Same Sport
The total tells a family how much capacity exists. It does not tell them how that capacity is arranged, and the arrangement is what determines strategy. A sport can be available everywhere and offer almost nothing at each stop, or exist at a third of the colleges and offer a great deal at each.
Share of colleges sponsoring each sport, plotted against average roster places per entering class at the colleges that sponsor it. Bubble size shows total places across all 26 colleges. Data: EADA sport-level file, 2024-25.
Only one sport sits in the deep and wide corner. Track and cross country is sponsored by all 26 colleges and carries about 30 adjusted places per entering class at each, ahead of football at 21.7 and rowing at 19.6, and unlike either of those it exists everywhere. Nothing else combines both.
Rowing and ice hockey occupy the deep and narrow quadrant. Each exists at 9 of the 26 colleges and each carries a substantial squad, 19.6 and 13.2 places per class respectively. For a recruit in either sport, the college list is effectively chosen for them, and the schools that do sponsor the sport have real capacity to fill.
The shallow and wide quadrant is the most crowded and the most misleading. Basketball, tennis, volleyball, golf and softball are sponsored by nearly every college, which makes them look like the safest sports to be recruited in. Each carries between 4.4 and 7.7 places per entering class. Basketball, sponsored by all 26, accounts for 7.7 places per class at each, which across two rosters means a coach is filling roughly four positions per side per year.
Lacrosse and football sit alone in the middle, sponsored by roughly three-quarters of the colleges with squads large enough to matter. Lacrosse is the more useful of the two for most families, because it exists on both the men’s and women’s side at 19 colleges and carries 19.4 places per class.
What This Analysis Cannot Tell You
Placing the limitations here rather than at the end is deliberate, because the next section is the one most likely to be misread.
Roster size and recruiting intensity are not the same variable, and they diverge most in exactly the sports that look largest here. Track and cross country carries large rosters partly because distance running accommodates unrecruited students who walk on and stay. A 30-place figure does not mean 30 recruited runners. Squash, at 6.7 places per class, is probably closer to fully recruited than track is, despite being an eighth the size.
The data also says nothing about competitive standard. Two colleges sponsoring the same sport with the same squad size can want entirely different athletes, and a recruit who is supported at one may not draw a conversation at the other.
The duplication figures are estimates produced by an allocation assumption, not corrected counts. Public data records the total number of distinct athletes at each college and the number of participants in each sport, but never which sports the duplicated athletes played. Students who play two different sports, football and track or field hockey and lacrosse, are duplicated in the same way a multi-season runner is, and this analysis has no way to separate them. The true figure for track sits somewhere between 789 and 1,396 places per entering class. Treat 789 as a floor rather than an answer. Nothing here distinguishes a cross country recruit from a throws recruit either, and those are different admissions conversations.
Where the Openings Are on the Women’s Side
Football creates a large participation imbalance on the men’s side at the 20 colleges that field it. About 435 men’s roster places per entering class sit in a sport with no women’s equivalent, which helps explain why colleges with football tend to maintain substantial participation opportunities across women’s sports. Federal rules do not require a college to offset each football place with a women’s one; institutions can meet the participation standard through substantial proportionality, a history of expanding opportunities, or full accommodation of the underrepresented sex’s interests.
Estimated roster places per entering class by sport and sex, summed across 26 elite Division III colleges. Sports accounting for fewer than 40 places per class are omitted. Data: EADA sport-level file, 2024-25.
Field hockey and softball are women’s sports at every college that sponsors them, accounting for 109 and 107 places per entering class. Volleyball runs 114 women against 15 men. Swimming and diving is the one large sport where women hold more places than men, about 114 against 108 on the adjusted figures.
Field hockey deserves particular attention from families outside the Northeast. Eighteen of the 26 colleges sponsor it, almost all of them in the Northeast and mid-Atlantic, and it is a sport many high schools elsewhere do not offer at all. That combination means the supply of recruitable players is geographically concentrated in a way that few other sports are.
On the men’s side, the concentration runs the other way. Football, baseball and wrestling account for 709 places per entering class in sports with no women’s counterpart. A male recruit in one of those three is competing in a larger market than the sport’s visibility suggests.
How to Use This When Building a List
Start by finding out how many of the target colleges sponsor the sport at all. For rowing, ice hockey, squash, wrestling, fencing and skiing, the answer is fewer than 15 of 26, and for several it is fewer than 10. That single fact reshapes a college list more than anything else in this analysis.
Then check squad size at those specific colleges rather than the national figure. A sport averaging 5 places per entering class across 26 colleges may carry 9 at one of them and 2 at another, and the difference decides whether a coach has room in a given cycle.
Treat the wide and shallow sports with more caution than their availability suggests. Being recruitable in basketball or golf means almost every college on the list is theoretically in play, and it also means every one of those coaches is filling a handful of positions and can afford to be selective. Breadth of opportunity and depth of roster capacity often move in opposite directions.
For counselors: the sports that most reward early identification are the deep and narrow ones. A student who might row has a nine-college universe, and if that student and family do not know it by junior year, the list gets built around the wrong seventeen schools. The same logic applies in reverse for track, where the capacity is real at every college but the walk-on culture means a coach’s support is a different thing from a roster place. Ask what the coach is offering, in those words.
Ask coaches how many positions they are filling this cycle and how many they expect to support in admissions. Those are two different numbers and the gap between them is the whole story for any individual student.
The numbers here narrow a list from twenty-six colleges to the handful where a given sport has real capacity. Working out which of those coaches will actually spend a support position on a particular student is the next conversation, and it is one worth having before junior year ends.
Methodology
Sport-level participation comes from the Equity in Athletics Disclosure Act sport file for the 2024-25 academic year, the most recent available. The universe is the 26 Division III colleges admitting 35 percent or fewer applicants that appear in the selective college set used in the Chetty, Deming and Friedman pipeline data, with the Claremont consortium colleges excluded because Pomona-Pitzer and Claremont-Mudd-Scripps field joint teams that cannot be attributed to a single college.
EADA assigns each sport a numeric code, and two of those codes are aggregates. Code 3 combines track and field with cross country, and code 18 combines swimming with diving. Colleges report either the combined category or its components and never both, which we verified across the full universe. Fourteen colleges use the combined track category and twelve report components separately. We merge indoor track, outdoor track and cross country into a single category, and swimming with diving into another, so that all 26 colleges are measured on the same basis.
Summing participants across sports counts athletes more than once whenever they appear on more than one roster. Across these 26 colleges the sport file contains 18,141 roster entries against 15,059 distinct athletes in the institution-level file, a gap of 3,082, or 20.5 percent. Wellesley College files an institution-level report but no sport-level report for 2024-25, so it is excluded from this article although it appeared in the previous one.
The estimates in this article allocate that entire gap to track and cross country and to swimming and diving. That is an assumption, not a measurement. A student who plays soccer in the fall and lacrosse in the spring also appears twice in the sport file and once in the unduplicated count, and public data does not record which sports any duplicated athlete played. Allocating everything to the two multi-season categories therefore pushes the track and swimming figures to the low end of their plausible range and leaves every other sport slightly overstated. We report it this way because multi-season participation within a single track or swimming program is structural and near-universal, while cross-sport participation is occasional. Charts show both the estimated and the unadjusted values for the two affected sports.
Roster places per entering class divide total participants by four class years. This assumes even distribution across class years, which approximates but does not reproduce the composition of any single cycle. Sports sponsored by fewer than three colleges are omitted from the table and charts.
Participation counts include walk-ons and students who joined a team after enrolling. No public dataset separates recruited athletes from other participants at the sport level, which is why this analysis describes capacity rather than recruiting activity.
Data: Equity in Athletics Disclosure Act sport-level and institution-level files, 2024-25, U.S. Department of Education. IPEDS admissions component, fall 2024, used to define the selective universe. Statistics on the College Pipeline, Opportunity Insights (Chetty, Deming and Friedman, 2023). Roster places are calculated as total participants divided by four and include walk-ons. Charts: College Transitions Data Team.
