The testing effect refers to the finding that retrieval of previously learned information improves retention of that information more than restudy practice does. While there is some evidence that the testing effect can already arise in preschool children when a particular experimental task is employed, it remains unclear whether, for this age group, the effect exists across a wider range of tasks. To examine the issue, the present experiments sought to determine the potential roles of retrieval-practice and final-test formats, and of immediate feedback during retrieval practice for the testing effect in preschoolers. Experiments 1 and 2 showed no testing effect in preschoolers when a free-recall task was applied during the final test, regardless of whether free recall (Experiment 1) or cued recall (Experiment 2) were conducted during retrieval practice. In contrast, if cued-recall tasks were used during both retrieval practice and the final test (Experiment 3), a reliable testing effect arose. Furthermore, the magnitude of the effect was dramatically enhanced when, in addition, immediate feedback was provided during retrieval practice (Experiment 4). The present findings suggest that cued-recall practice and test formats, as well as immediate feedback during practice, are crucial ingredients for obtaining the testing effect in preschoolers.