The last Super Bowl was Super Bowl LIX, played February 9, 2025, at Caesars Superdome in New Orleans. The Philadelphia Eagles defeated the Kansas City Chiefs 40–22. The Eagles' active roster and practice-squad call-ups, plus the Chiefs' full 53-man roster and game-day elevations, are the players who played in the last Super Bowl, with 46 players per team eligible to dress on game day.

That answer is simple. The interesting science sits underneath it: how a 53-man roster becomes 46 dressed players, why snap counts distribute the way they do, and why "played" can mean anything from 71 offensive snaps to a single special-teams rep. This explainer goes one level deeper than a box score.

Roster mechanics: how 53 becomes 46 on game day

An NFL team carries 53 players on its active roster during the season. On game day, only 46 can dress. That 7-player gap is the first filter on who actually plays in a Super Bowl, and it is a coaching decision made roughly 90 minutes before kickoff, when inactive lists are submitted.

The mechanism behind the number is a collective bargaining rule: teams may dress 48 players if at least 8 are offensive linemen, and may add two practice-squad elevations to reach 46, subject to elevation limits. Practice-squad players elevated for a game revert to the squad afterward, which is why some Super Bowl participants never appear on a 53-man roster at all.

Dressing is not the same as playing. A dressed backup quarterback may take zero snaps. A dressed cornerback may play 40 special-teams snaps. The distinction matters because "who played" in a literal sense includes anyone who stepped on the field for at least one snap, and that set is smaller than the dressed set.

Injuries compress the margin further. A team that loses a lineman mid-game may finish with fewer than 46 available players, forcing defensive players onto special teams and offensive players into emergency roles. That cascade is one reason Super Bowl rosters are often described by position groups rather than names.

Plain language: the league lets 53 players be on the team, but only 46 can put on pads, and fewer still touch the field.

Quantified context: snaps, participation, and the 2025 game

Snap counts are the cleanest measure of who played. In Super Bowl LIX, the Eagles ran 71 offensive snaps and the Chiefs ran 61, and defensive snap counts mirror the opponent's offensive total. A player with a 100% snap share on offense therefore logged 71 reps, while a rotational defensive lineman might log 25 to 35.

Special teams add a separate ledger. A core special-teamer can appear on 20 to 25 kicking-game snaps without registering an offensive or defensive rep. This is how a player with a tiny box-score line still "played in the Super Bowl" in a way that matters to coaches and contract incentives.

Participation data is compiled from game film and official play-by-play, not from the box score. The box score lists statistics; participation charts list who was on the field. Those two records diverge constantly, especially for offensive linemen, who record no statistics at all yet play every snap.

Here is the practical threshold structure, reconciled:

Measure Threshold to count as "played" Source of truth
Active roster 53 players Team roster filing
Dressed 46 players (48 with 8+ OL) Inactive list, ~90 min pregame
Practice-squad elevation Up to 2 per game, limited elevations Transaction wire
Participation 1+ snap on offense, defense, or special teams Participation chart
Statistical appearance 1+ recorded stat Official box score

If you want a defensible answer to "who played," use the participation chart. If you want the roster answer, use the 53-man list plus elevations. Both are correct under different definitions, which is why thin explainers often contradict each other.

Edge cases: inactive stars, elevation limits, and the injury cascade

The most counterintuitive detail in Super Bowl roster science is that a healthy star can be inactive. Coaches sometimes deactivate a veteran to protect a younger player's development path, or to keep an extra lineman for protection. Being on the 53 does not guarantee dressing, and dressing does not guarantee a snap.

Practice-squad elevations create a second edge case. A player elevated for the Super Bowl is on the game-day roster, can play, and then returns to the practice squad. These players are legitimately Super Bowl participants, but they may not appear on the team's official 53-man roster page the next morning.

Injury designations add a third layer. Questionable, doubtful, and out designations are probability labels, not binary states. A player listed as questionable can dress and play 50 snaps, or dress and play none. The designation reflects medical risk, not a commitment.

Red flag: treating a box score as a complete participation list. Red flag: assuming a 53-man roster spot means a player dressed. Red flag: assuming a practice-squad elevation is not a real Super Bowl appearance. Red flag: counting a player who was inactive as having "played." Red flag: conflating special-teams-only reps with no contribution.

The injury cascade is the mechanism that ties these together. When a starter goes down, the next man up inherits snaps, and a previously inactive player may become active in subsequent weeks. In a single game, the cascade is immediate: a torn ligament can move a backup into 40 snaps he never practiced for.

Misconceptions: what "played in the Super Bowl" does not mean

The largest misconception is that the Super Bowl roster equals the season roster. It does not. Trades, waivers, injuries, and elevations mean the players who suited up in February may differ substantially from the Week 1 roster. A fan searching for a favorite player may find he was traded in October.

A second misconception is that all dressed players see the field. Coaches routinely dress a third quarterback or a developmental lineman as insurance and never call their number. Those players are Super Bowl participants in the roster sense but not the participation sense.

A third misconception is that participation charts are official league statistics. They are compiled by teams and independent charting services, and they can disagree at the margins on whether a player was on the field for a given snap. The league does not publish a single canonical participation dataset.

A fourth misconception is that special teams do not count. They do. A gunner who covers four punts played in the Super Bowl, and his reps are charted. The box score simply does not reflect them, which is why thin explainers miss him entirely.

Plain language: the roster tells you who was eligible, the participation chart tells you who actually moved, and the box score tells you almost nothing about either.

Timeline: from roster filing to final whistle

The sequence that determines who plays in a Super Bowl runs over roughly two weeks, and each stage narrows the pool:

When Event Effect on who plays
~2 weeks before Conference championships end Rosters frozen except for elevations and injury replacements
Super Bowl week Practice squad elevations submitted Up to 2 players added to game-day pool
~90 min pregame Inactive lists submitted 7 players (or 5 with 8+ OL) excluded from dressing
Kickoff 46 dressed players Participation begins
In-game Injuries, ejections Snap redistribution to backups
Final whistle Participation chart compiled Definitive "who played" record
Postgame Elevations revert Practice-squad players return to squad

The 90-minute inactive list is the single most consequential gate. It is submitted after warmups, so a player who tweaks a hamstring in pregame warmups can be scratched at the last moment, and a previously inactive player can be activated in his place. This is why final rosters are not known until shortly before kickoff.

The postgame reversion is the quiet edge case. An elevated player who played 20 snaps in the Super Bowl can be back on the practice squad the next week, and his Super Bowl appearance remains real even though his roster status changed.

Practical actions: how to verify who played, and how to read the data

If you need a defensible list of who played in the last Super Bowl, follow a fixed procedure rather than trusting a single source. The steps below reconcile the thresholds from the table above.

Start with the official game book and the two teams' 53-man rosters as of game week. Add practice-squad elevations from the transaction wire for that week. Pull the inactive list submitted roughly 90 minutes before kickoff. Cross-reference the participation chart for snap counts by unit. Treat any player with 1+ charted snap as having played. Note special-teams-only players separately, because the box score omits them.

For a quick answer, the 46 dressed players per team plus elevations is the roster-level answer. For a precise answer, the participation chart is the participation-level answer. For a statistical answer, the box score is the narrowest and least complete.

The threshold reconciliation is explicit: 53 is eligibility, 46 is dressing, 1 snap is participation, and 1 stat is a statistical appearance. Each answers a different question, and mixing them produces the contradictions common in thin explainers.

Limitations: what participation data cannot tell you

Participation charts do not measure effectiveness. A player can log 60 snaps and grade poorly, or 15 snaps and change the game. Snap counts describe exposure, not impact, and they are compiled by human charters who can miss a rep at the margins.

The data also cannot resolve intent. A coach may dress a player as a decoy, a mentor, or a contingency, and none of those roles appear in a chart. The medical designations that shaped the inactive list are probability labels, not diagnoses, and they are not public in detail.

Finally, the record is retrospective. The last Super Bowl's participation data is stable now, but the next one will be determined by a new set of inactives, elevations, and injuries that cannot be predicted from the prior game. Any claim about who will play in a future Super Bowl is a projection, not a fact.

This article is general educational information about NFL roster and participation mechanics, not a prediction, betting recommendation, or medical assessment. It does not promise outcomes or returns. For decisions with financial or health consequences, consult a qualified professional.