Experimental studies of dynamic decisionmaking generally showpoor performance.Most, however, lack market mechanisms, specifically price setting, while economic theory suggests markets should mitigate individual decision errors. We develop experimental markets to explore whether different price institutions improve performance in dynamic decision tasks.We find: (i) dynamic complexity degrades performance substantially relative to optimal despite the inclusion of different pricing mechanisms; and (ii) markets improve performance, though it remains significantly below optimal. We estimate decision rules for each actor; results reject the hypothesis of rationality at the individual level but support behavioral decision rules consistent with bounded rationality. Simulations using the estimated decision rules reproduce key features of market dynamics. Decision timing data and verbal protocols show that greater task complexity leads subjects to ignore important aspects of the environment, particularly strategic interactions among participants. Markets moderate but do not eliminate misperceptions of feedback.