The present work analyses the predictive validity of measures provided by several available self-report and indirect measurement instruments to assess risk propensity (RP) and proposes a measurement instrument using the Implicit Association Test: the IAT of Risk Propensity Self-Concept (IAT-RPSC), an adaptation of the prior IAT-RP of Dislich et al. Study 1 analysed the relationship between IAT-RPSC scores and several RP self-report measures. Participants’ risk-taking behaviour in a natural setting was also assessed, analyzing the predictive validity of the IAT-RPSC scores on risk-taking behaviour compared with the self-report measures. Study 2 analysed the predictive validity of the IAT-RPSC scores in comparison with other indirect measures. Results of these studies showed that the IAT-RPSC scores exhibited good reliability and were positively correlated to several self-report and indirect measures, providing evidence for convergent validity. Most importantly, the IAT-RPSC scores predicted risk-taking behaviour in a natural setting with real consequences above and beyond all other self-report and indirect measures analysed.