The Preliminary Scholastic Aptitude Test/National Merit Scholarship Qualifying Test (PSAT/NMSQT) was equated to the Scholastic Aptitude Test (SAT) using item response theory (IRT) in a three-parameter logistic model. Estimated true formula score equating was used. Only the mathematical sections of the PSAT/NMSQT and SAT were used because the discrepancy in level of difficulty is greater than in the verbal sections and is most representative of a vertical equating situation. The equating used an internal anchor test design. The results of the IRT equating method were compared to those obtained from the conventional linear and curvilinear equating methods. The use of a goodness of fit statistic leading to a chi-square-like test in conjunction with item ability regression plots is examined. The lack of a criterion for judging the equatings and the subjective nature of the goodness of fit assessment are discussed. (CM)