Piscine venom glands have implicitly been assumed to be anti-predatory adaptations, but direct examinations of the potential fitness benefits provided by these structures are relatively sparse. In previous experiments examining this question, alternative phenotypes have not been presented to ecologically relevant predators, and the results are thus potentially confounded by the presence of sharp, bony fin spines in these species, which may also represent significant deterrents to predation. Here, I present the results of experiments exposing Micropterus salmoides (largemouth bass) to tadpole madtoms (Noturus gyrinus) with one of several fin spine phenotypes (intact, stripped, absent), which indicate that the venom glands of this species do provide a significant fitness benefit, relative to individuals having fin spines without venom glands or no spines at all. Intact madtoms were repeatedly rejected by the bass and were almost never consumed, while alternative phenotypes were always consumed. Madtoms with stripped fin spines showed increases in predator rejections relative to spineless madtoms and control minnows, but non-significant increases in handling time, contrasting with previous results and predictions regarding the adaptive benefit of these structures. Comparisons with a less venomous catfish species (Ameiurus natalis) indicate that a single protein present in the venom of N. gyrinus may be responsible for providing the significant selective advantage observed in this species. These results, considered in conjunction with other studies of ictalurid biology, suggest that venom evolution in these species is subject to a complex interplay between predator behavior, phylogenetic history, life history strategy and adaptive responses to different predatory regimes.