A national sample of 46 nurse practitioners (NPs) and a Midwestern sample of 31 nurses completed five instruments designed to provide evidence of the construct validity of patient management problems (PMPs). The subjects completed three simulations, a multiple-choice cognitive examination in the content area of the simulations, and a demographic questionaire. The first hypothesis stated that a claim for the construct validity of simulations would be warranted if significant differences were found between performance of NPs and nurses on simulations designed for NPs. A second hypothesis stated that no significant differences would be found across the three different simulations. The cognitive examination was hypothesized to be a meaningful covariate, as performance on a simulation depends in part on knowledge of the content area of the simulations. The examination was not found to be a significant covariate in this analysis. Meaningful differences were found between NPs and nurses, supporting the claim for the construct validity of the simulations. Meaningful differences were also found across the three simulations for both groups of subjects, suggesting that the second hypothesis of consistency of performance may have been inappropriately formulated. One measurement characteristic of the simulations may be difficulty level.