The current study extends existing research on human error by investigating human error in complex, dynamic, and uncertain microworlds. The main goal was to develop a taxonomy of errors for such situations. Several tasks with differing characteristics and demands on the problem solver were used: the simulation of a chocolate producing company (CHOCO FINE), the simulation of a tribe of semi-nomads in the Sahel zone (MORO), and the simulation of wildfires engaging fire-fighting units (WINFIRE). Observing participants and teams working on these simulations, we refer to specific cases and describe 24 errors related to the steps of human problem solving and dynamic decision making. A theory is then presented which attempts to explain the causes of these errors occurring in microworld contexts. Whereas most existing theories focus on cognitive explanations, the theory presented explains human error as a result of the interaction of emotions, motivations, and cognition. The current study not only extends the taxonomy and list of human errors, but especially addresses the lack of theoretical explanations in the literature on human error. The findings provide a starting point for further theory development and empirical testing and could be applied in training programs; hopefully contributing to sensitivity and awareness of the demands of complex, dynamic, and uncertain problems, and ultimately contributing to fewer human errors.