Replicated E. Tulving's (see record) study which used a paired-associate procedure in which "B" terms were recalled better than they were recognized. 20 undergraduates were presented with and tested on a practice list and then shown an experimental list of which certain pairs had been designated as critical. All "A" terms had 5 letters and all B terms, 2 letters. Immediate testing was done on noncritical pairs. After a filled interval, all Ss were given free and forced-choice recognition tests of the critical B terms, and finally, critical A terms were presented in a conventional cued-recall test of the corresponding B terms (e.g., critical terms were SPANI - EL or EXPLO - RE). Results replicated Tulving's finding; for all Ss, number of presented critical B terms successfully recalled was greater than the number freely recognized and equalled or exceeded the number responded to in the forced-choice recognition test. It is suggested that results do not support current models in which recall and recognition are viewed as distinct processes. (PsycINFO Database Record (c) 2006 APA, all rights reserved)