Most dynamic decision making tasks include assumptions which have a huge uncertainty attached to them. Organizations are inherently complex. The combination of uncertainty and complexity results often in a sub-optimal decision. This paper emphasises on the usage of probabilistic system dynamics (SD). The focus of probabilistic SD is to represent the behaviour of uncertain variables in a realistic manner. The information generated by probabilistic SD could produce “complete” information thereby improving the mental models of decision makers. Many SD models use deterministic values of variables. However, “determinism” is untrue for real business settings. In order to test the effectiveness of probabilistic SD on managerial decision making, this study aims at conducting a series of rigorous and controlled experiments. Specifically it tests the usefulness of (1) system dynamics itself, (2) model validation techniques and (3) probabilistic system dynamics on decision-making. Furthermore, these experiments are conducted in two settings – (1) using a simple model and (2) using a complex model. It is hoped that probabilistic SD would be instrumental in producing relevant information that would help in improving managers’ mental models, especially in complex scenarios. This in turn will result in better decisions under uncertainty in complex business environments.