This study examines whether National Football League (NFL) teams allocate salary cap resources efficiently, where an efficient allocation is one that maximizes a team's probability of winning. Using team- and player-level data from the 2018 through 2025 NFL seasons (256 team-seasons; 13,604 player-seasons), teams are classified into three groups by regular-season win total: Rebuilding (0-6 wins), Competitive (7-10 wins), and Contender (11+ wins). Efficiency is tested two ways. First, an ordinary least squares regression estimates the association between wins and the percentage of salary cap allocated to each of fourteen position groups; similar coefficients across positions would indicate efficient allocation across a team's roster. Second, at the player level, the correlation between a player's cap share and his individual Expected Points Added (EPA) is examined by team category, since efficient allocation predicts that more productive players receive a larger share of team cap. A performance-per-dollar metric is also constructed to identify which positions generate the most on-field value relative to their cost. Results show that quarterback, wide receiver, and offensive line spending are most closely tied to wins, while running back and linebacker spending show no significant association with wins or performance. The pay-to-production relationship is effectively absent among Rebuilding teams (r = 0.008) but moderately strong among Contenders (r = 0.481), indicating that contending teams allocate cap more efficiently at the player level than rebuilding teams do. Cap efficiency and season-to-season consistency were found to be largely independent traits, meaning the most cap-efficient teams are not necessarily the most consistent from year to year. These findings suggest that NFL teams at different stages of contention would benefit from different salary cap strategies, with quarterback spending representing the clearest point of divergence between winning and losing teams.