Interest in fluctuating asymmetries, random deviations from perfect bilateral symmetry, has spread to studies of sexual selection because of the intriguing idea that females could use the degree of asymmetry of a male trait to assess the genetic quality of potential mates. The evidence that females prefer males with symmetrical sexual signals, however, remains controversial. A problem that applies to most previous studies is that preference for trait size can be misinterpreted as preference for symmetry, even when overall trait size is held constant, if females assess trait size by the largest minimum on one side. If overall trait size is equal between males, the asymmetrical males will have the maximum and minimum trait size, and so preference to mate with symmetrical males could actually reflect a preference to avoid males with the minimum trait on one side. Xiphophorus cortezi females preferred males with symmetrical bar numbers when the minimum number of bars was held constant. The strength of female preference for the symmetrical males was negatively correlated with the strength of preference the same females had for bar number. These results clearly demonstrate that females preferred trait symmetry in addition to trait size.