In the studies of decision making under risk and time, an increasingly important question is whether the utility representations should be the same for both circumstances. This study aims to test the separation between risk preference and time preference in a controlled laboratory setting, where subjects make intertemporal allocation decisions under different types of correlated risks. Our results show that risk and time preferences are distinct, and furthermore support the notion proposed by Epstein and Zin (1989) on the separation between risk attitude and intertemporal substitution. In addition, the estimated parameter for risk attitude is significantly smaller than that for intertemporal substitution. These findings shed light on recent literature linking risk preference and time preference.