The maintenance of genetic variation in traits under strong sexual selection is a longstanding problem in evolutionary biology. The genic capture model proposes that this problem can be explained by the evolution of condition dependence in exaggerated male traits. We tested the predictions that condition dependence should be more pronounced in male sexual traits and that genetic variance in expression of these traits should increase under stress as among-genotype variation in overall condition is exposed. Genetic variance in female and nonsexual traits should, by contrast, be similar across environments as a result of stabilizing selection on trait expression. The relationship between the degree of sexual dimorphism, condition dependence and additive genetic variance (Va) was assessed for two morphological traits (body size and relative fore femur width) affecting male mating success in the black scavenger fly Sepsis punctum (Diptera: Sepsidae) and for development time (a nonsexual trait often correlated with body size). We compared trait expression between the sexes for two cross-continental populations that differ in degree of sexual dimorphism (Ottawa and Zurich). Condition dependence was indeed most pronounced in males of the strongly dimorphic Zurich population (males larger), and Va was similar for males and females unless the trait was strongly sex specific and condition dependent. Contrary to prediction, however, Va primarily increased under food limitation in both sexes, and genetic variance in fore femur width was low to nil, perhaps depleted by putatively strong sexual selection. Solely for body size of Zurich males, Va increased more in males than females at limited food, in accordance with the predictions of the genic capture model. Overall therefore, quantitative genetic evidence in support of the model was inconsistent and weak at best.