Performing a pretest (plate - ?) before studying the material to be learned (plate - fork) can improve long-term retention of the material compared to material that was only studied initially. In the present study, weakly associated word pairs were used to investigate whether this pretest effect is modulated in magnitude when pretests are repeatedly administered during acquisition. To this end, two experiments were conducted in which subjects were presented with word pairs that could either be studied immediately or were given up to three pretests. Recall performance on a later final test consistently showed the typical pretesting effect, with improved recall after a single guessing attempt compared to the study-only baseline. Crucially, the pretesting effect increased in size when multiple guessing attempts were made during acquisition, irrespective of whether the length of the pretesting phase increased with the number of guessing attempts (Experiment 1) or was held constant (Experiment 2). Overall, the results show that additional guessing attempts can promote access to the pretested material on the final test, indicating that in educational contexts, extended pretests during learning may provide an effective learning tool.