Sampling biases resulting from capture methodology or animal behavioral responses can significantly skew our understanding of population size and structure. Behavioral responses to passive sampling with traps, such as trap-happy or trap-shy responses, can be accounted for in analyses using individual covariates but they require detailed understanding of the underlying behavioral mechanisms. We used a field experiment to investigate the impact of trap escape on capture rates and determine what traits influence trap escape rates between and within two species of semiaquatic snakes, Liodytes pygaea and Nerodia fasciata. We found that L. pygaea escaped significantly more often than did N. fasciata, smaller L. pygaea escaped at significantly higher rates than did larger conspecifics, and individual capture history significantly influenced escape rates in N. fasciata. Our findings highlight understudied sources of inter- and intraspecific capture heterogeneity in a common sampling technique, and we urge researchers to consider method-specific sampling biases when attempting to produce population parameter estimates for species that are difficult to detect.