Fluctuating asymmetry (FA) may provide a direct test of whether sexual selection via “arbitrary traits” or “good genes” is responsible for the evolution of epigamic traits, because FA is an epigenetic measure of stress during development. However, tests of the FA hypothesis have yielded equivocal results, and the debate between arbitrary traits and good genes continues. The FA hypothesis predicts a negative relationship between ornament size and asymmetry. In non-ornamental traits, a U-shaped relationship between trait size and asymmetry is expected. We tested these predictions in the Red-collared Widowbird (Euplectes ardens) by examining the relationship between size and asymmetry in the length of the tail, tarsus, and wing. We found no significant linear or second-order polynomial relationships between trait size and asymmetry. Furthermore, no relationship existed between asymmetry and two indices of body condition. This suggests that the tail is not a reliable signal of male quality in the Red-collared Widowbird. However, we argue that the assumptions of the FA hypothesis are too simplistic for this conclusion to be drawn with a high degree of confidence.