<p>In this study, we examine the correlation between religiosity and risk taking among the Muslim population. We elicit risk preferences of Muslim subjects in Pakistan in an incentivized laboratory experiment and a non-incentivized field-survey. We operationalize religiosity by using belief-based, ritual-based, and knowledge-based items. We find robust evidence that very religious people as measured by the frequency of prayers (a ritual-based measure) take less risk than less religious people. In contrast, we find inconclusive evidence for the influence of religiosity on risk taking when elicited through other ritual-based measures and measures based on beliefs or knowledge. Our findings suggest that religiosity and risk-taking nexus among Muslims is sensitive to the religiosity measures. Therefore, we discuss the relative effectiveness of different religiosity measures and highlight why the frequency of prayers can be a useful proxy for measuring the religiosity of Muslims.</p>