Florida queen butterflies are highly variable in cardenolide content and, in three populations studied, contained less cardenolide than did a sample of sympatric Florida monarchs. The possibility that queens stored a more potent set of cardenolides from their host plants (and therefore were as well protected as monarchs, even at lower concentrations) is refuted by Chromatographic analysis of wild butterflies, as well as controlled laboratory rearings. It therefore appears that, with respect to cardenolides, monarchs are better defended than are queens. Consequently, cardenolides are unlikely to explain the apparent shift in Florida viceroy mimicry away from resemblance of the monarch, toward mimicry of the queen. Other hypotheses to explain this mimetic phenomenon are suggested. Adult monarchs exhibit significant negative correlations between the concentration of cardenolide stored in their tissues and both body size and weight, whereas queens show no such correlations. The implications of these results for the study of "metabolic costs" of allelochemic storage are discussed. Chromatographic evidence is provided that monarchs do breed in south Florida during the winter months and that the likely host plant employed by the population studied wasAsclepias curassavica. This represents the first practical application of cardenolide "fingerprinting" to identify the larval host plants of wild danaid butterflies.