Can risk-taking propensity be thought of as a trait that captures stable individual differences across diverse scenarios and time? Or is it best conceptualised as a domain- or situation-specific behavior? I present cross-sectional and longitudinal studies investigating individual and age differences in risk-taking propensity and show that while there are some clear patterns of stability and change across the life span in self-reported risk taking, there is less coherence and little convergent validity across behavioral measures. I discuss the promises and challenges posed by such results for the investigation of the neural basis of individual and age differences in risk taking.