The effects of varying degrees of correlation between abilities and of various correlation configurations between item parameters on ability and item parameter estimation using the three parameter logistic model were examined. Ten two-trait configurations and one unidimensional test configuration for 30 item tests were simulated. Each configuration consisted of a specific item parameter configuration and a specific correlation between traits on two dimension. Six conditions were simulated for each configuration, ranging from an easy to a hard test. The accuracy of item and ability parameter estimation was examined using correlations; KR-20 coefficients and factor analyses were also performed. The factor analyses supported a division of the simulated multidimensional data sets into groups according to the discrimination parameter "loads" on the two dimensions. The tests either both load heavily on both dimensions (both tests are multidimensional), one test loads heavily on one dimension and the other loads heavily on the same dimension (both tests are unidimensional), one test is unidimensional and one is multidimensional, or one test loads heavily on one dimension and the other test loads heavily on the other dimension. Results indicated the poorest item parameter estimations occur for the situation in which one test is unidimensional and one is multidimensional. (Author/DWH)