The identity model proposes that anticipation and study-test methods of paired-associate learning differ little in basic acquisition, retrieval, and storage processes per event: study, test, and intervening study and test events of other items plus the intercycle interval. The major difference between the two methods boils down to differential short-term loss over the differential retention interval. When the study-test interval is equated, the identity model predicts no difference between the methods. However, the feedback model predicts superiority for the anticipation method, and the differential acquisition, context effect, or task alternation models predict superiority for the study-test method. An experimental test utilized (nearly) identical study-test intervals, various materials, and both massed and spaced practice. In 15 out of 16 comparisons, there was no difference between the methods, providing support for the identity model. Within limits, the greater the study-test interval differentials, the greater the advantage for the method with the shorter study-test intervals. Quantitative analyses support the identity model.