The purpose of this study was to examine the homogeneity in difficulty of item domains and the effectiveness of Rasch pre-equating procedures for adjusting test scores for differences in the difficulty of tests constructed by sampling from item domains. The data used were taken from a field test and calibration of 810 tenth-grade items in mathematics and communications skills. Approximately 250 students responded to each of the items. The mean range of item difficulty for 15 mathematics skills domains was 2.593 logits and for 10 communications skills domains was 4.210 logits. It was found that Rasch pre-equating techniques effectively adjusted total scores from tests of 45 and 53 items, selected from specific sub-domains, for substantial differences in the level of difficulty of items included in the tests. (Author/BW)