The efficacy of employing subsets of items from a calibrated item pool to estimate the Rasch model person parameters was investigated. Specifically, the degree of invariance of Rasch model ability-parameter estimates was examined across differing collections of simulated items. The ability-parameter estimates were obtained from a simulation of examinee responses to 36 item sets that differed both in fit to the model and in item difficulty. Three examinee groups were formed in order to vary only by shape of the distribution of true ability. The three item-subset characteristics were: (1) percentage of items in the subset that fit the model; (2) difficulty level of the items in the subset; and (3) degree of misfit of those items in the subset that do not fit the model. Each examinee's response to an item was simulated under a general logistic model which takes into account variation in item discrimination. The simulation resulted in an observed score for a subset of items which was based on the examinee's true ability, item difficulty, item discriminations, and a random-error component. (Author/CTM)