The males of numerous butterfly species fight with conspecific rivals to possess mating territories. Although there is little consensus on the nature of fight costs or on what traits favor victory, a recent analysis suggests that size may be of minor importance. However, data are inconsistent, and wing length, the metric that has been most widely used for expressing size in butterfly field studies, may not provide a reliable index of muscle mass or other traits that determine resource holding potential. Here we investigate the influence of wing length, body mass and wing wear on the territorial success of a neotropical satryine butterfly, Paryphthimoides phronius. In a removal experiment, original residents were heavier than replacement males that took over territories in the following minutes. Winners of naturally occurring contests were also heavier than losers after correcting for weight loss since weighing. Although wing length correlated significantly with body mass and body mass decreased with age, neither wing length nor wing wear (a surrogate for age and body condition) contributed to explaining territorial success. As body mass may be related to traits such as muscle mass, fat stores, and age, no conclusion is possible on how mass translates into success in disputes. However, because wing length is unrelated to residency status in P. phronius, whereas mass is both a reliable predictor and performs significantly better than wing length, our results may help explain the frequent failure of wing length to predict territorial success in butterfly field studies.