No normative data have been available for the recall of recently studied foreign-language vocabulary items. We report data from 200 university undergraduates who had three study-recall trials on Swahili-English translation equivalents (e.g. ardhi-soil). Performance on each of the three study-recall trials is reported for each of the 100 items. Recall varied from near-zero to near-perfect across the combinations of items and trials, thereby allowing investigators to choose combinations of items and trials that will produce any desired range of recall difficulty. Substantial improvements in recall of the items occurred across study-test trials, and therefore the items seem suitable both for investigations of the effect of acquisition variables and for most of the other goals that are satisfied by laboratory paired-associate items.