Females mated to attractive males are predicted to produce male-biased broods. Previous studies on zebra finches, Taeniopygia guttata, in which colored leg rings were used to alter male attractiveness, support this hypothesis. However, because molecular sexing techniques were not available, it was not known when during development this bias arose. Also, because both attractive (red-ringed) and unattractive (green-ringed) males were within the same aviary, assortative mating between treatments may have confounded the results. Using two different experimental designs, we tested whether the sex ratio of zebra finch eggs and chicks differed in response to paternal ring color whilst controlling for assortative mating between treatments. In the aviary experiment, birds could interact socially, but all males in an aviary bad the same leg ring color. In the cage experiment, each female was randomly assigned a red- or green-ringed mate, thus also eliminating assortative mating within treatments. Offspring were sexed based on plumage or using a molecular method. The sex ratio at laying did not differ between treatments in either the aviary (n = 313 eggs) or cage (n = 151 eggs) experiments, suggesting that female zebra finches do not manipulate the primary sex ratio in response to their mate's ring color. However, in the cage experiment we found greater male embryonic mortality, in the attractive group, Which resulted in a female-biased sex ratio at sexual maturity, that is, in the opposite direction to that found in previous studies. Possible explanations for the disparity between our results and those of previous studies are considered.