Extant research offers conflicting predictions about the effect of pay dispersion on team performance. We collected a unique dataset from the Italian soccer league to study the effect of intra-firm pay dispersion on team performance, under different definitions of what constitutes a "team". This peculiarity of our dataset can explain the conflicting evidence. Indeed, we also find positive, null, and negative effects of pay dispersion on team performance, using the same data but different definitions of team. Our results show that when the team is considered to consist of only the members who directly contribute to the outcome, high pay dispersion has a detrimental impact on team performance. Enlarging the definition of the team causes this effect to disappear or even change direction. Finally, we find that the detrimental effect of pay dispersion is due to worse individual performance, rather than a reduction of team cooperation.