Latent trait model calibration procedures were used on data obtained from a group testing program. The one-parameter model of Wright and Panchapakesan and the three-parameter logistic model of Wingersky, Wood, and Lord were selected for comparison. These models and their corresponding estimation procedures were compared, using actual and simulated test results, with respect to: (1) the ability to calibrate multivariate data, (2) the sample size needed for calibration, (3) the effects of item quality on the calibration, and (4) the operational costs. The one-parameter procedure provided equivalent ability estimates at a lower cost than the three-parameter procedure when basically unifactor group tests were used. This method was therefore recommended for item calibration of group administered, multiple choice tests. However, the goodness of fit of the models to the data definitely showed the three-parameter model to be superior. Whether the better fit will outweigh the higher costs can only be determined by further comparison of the usefulness of the ability estimates, in actual tailored testing. (Author/MV)