Zebra swallowtail (Eurytides marcellus) butterflies are stimulated to oviposit by a single compound, 3-caffeoyl-muco-quinic acid (1). Analysis of the aqueous extracts of the leaves of the host. Asimina triloba, showed that they contained stimulant 1, its isomer (2), and the flavonoids rutin (3) and nicotiflorine (4) as major components. We compared the concentrations of compounds 1-4 in terminal leaves (TL) and expanded leaves (EL) of the host plants at four different times throughout the growing season. In spring, the concentration of 1 was highest in TLs, and flavonoids were not detectable or present at low levels. As the season progressed, however, the concentrations of flavonoids increased, reached maxima by late summer, and then decreased as the plants started senescing. There were also significant differences in the concentrations of these compounds between TLs and ELs. In a choice assay with model leaves, we tested equivalent amounts of post-dichloromethane aqueous extracts made in spring (May) and in fall (September). September extracts received significantly fewer approaches and eggs. In greenhouse experiments with potted A. triloba plants, the butterflies chose some leaves to lay eggs, while others were rejected or ignored. Analyses showed that the concentrations of compound 1 were not significantly different in the three kinds of leaves. The flavonoids (3 and 4), however, were significantly higher in the leaves that were ignored. Multiple-choice tests using model plants suggested that concentrations of both flavonoids and stimulant were important in assessing host suitability. There was a gradual decrease in approaches as the concentration of 1 decreased. Higher amounts of flavonoids deterred egg laying even in the presence of high concentrations of stimulant 1. At lower concentrations of 1, the addition of low doses of flavonoids deterred egg laying. Thus, the results suggest that the butterflies use both qualitative and quantitative information about these compounds to assess host quality. This behavior may have evolved to take advantage of seasonal variation in the chemistry of their host, A. triloba.