Fish behaviors have great potential as models for the study of pharmacology, genetics, and neuroscience. Zebrafish (Danio rerio), Japanese medaka (Oryzias latipes) and Chinese rare minnow (Gobiocypris rarus) are popular freshwater animal models. However, their behavioral use in aquatic toxicology research is generally hampered by oversimplified behavioral tasks and the fact that they are not well-developed animal models for toxicology. Here, this study presented a comparative analysis of multiple behavioral traits (i.e., anxiety-like behavior, novel object recognition, social preferences, habituation to light-dark stimulus and noise stimulus, and spatial learning and memory). We found that only medaka (d-rR) presented a weak or no response to repeated light-dark stimulus and noise stimulus. In addition, no significant behavioral changes were observed for the three species of juvenile fish models after 7 days of exposure to 0.01% v/v carrier solvents (i.e., ethanol, acetone, and DMSO). In contrast to zebrafish and Chinese rare minnow, medaka showed no significant changes in spatial memory after subacute exposure to 1 mg/L imidacloprid or 2.5 μg/L chlorpyrifos (cpf); instead, a hyperactivity response in the open field test and reduced social time were induced by cpf and imidacloprid, respectively. Our results suggest that: (1) behavioral effects are negligible when using <0.01% v/v carrier solvents for behavioral assessment; (2) given the differences in sensitivities of behavioral responses, a single behavior used alone as an endpoint may be insufficient for estimating the toxic impacts of pesticides or other environmental contaminants. In conclusion, these results could have major implications for aquatic toxicology research and water quality monitoring and ecotoxicological risk assessment.