Health system costs in clinical laboratories are incurred daily due to human error. Indeed, a major impetus for automating clinical laboratories has always been the opportunity it presents to simultaneously reduce cost and improve quality of operations by decreasing human error. But merely automating these processes is not enough. To the extent that introduction of these systems results in operators having less practice in dealing with unexpected events or becoming deskilled in problemsolving, however new kinds of error will likely appear. Clinical laboratories could potentially benefit by integrating findings on human error from modern behavioral science into their operations. Fully understanding human error requires a deep understanding of human information processing and cognition. Predicting and preventing negative consequences requires application of this understanding to laboratory operations. Although the occurrence of a particular error at a particular instant cannot be absolutely prevented, human error rates can be reduced. The following principles are key: an understanding of the process of learning in relation to error; understanding the origin of errors since this knowledge can be used to reduce their occurrence; optimal systems should be forgiving to the operator by absorbing errors, at least for a time; although much is known by industrial psychologists about how to write operating procedures and instructions in ways that reduce the probability of error, this expertise is hardly ever put to use in the laboratory; and a feedback mechanism must be designed into the system that enables the operator to recognize in real time that an error has occurred.