Prior accounting education research claims learning outcomes are improved by grouping together similar accounting practice problems rather than presenting such problems in an interleaved order. The present study revisits this prior research by asking whether making initial problem solving easier inadvertently leads to less durable longer-term learning. The evidence in the present study confirms that grouping practice problems helps students complete problem-solving practice in less time and with greater accuracy; this performance improvement is evident on a test given immediately after problem-solving practice. However, grouping together similar practice problems significantly reduces longer-term learning, as measured by a delayed test given one week after problem-solving practice. Further, the present study shows the efficient problem-solving experience created through grouping practice problems fools students into thinking they will be able to successfully solve similar problems in the future, and it also misguides them into believing they will need to study less when preparing for an upcoming test involving similar problems. This study raises the possibility that initial instruction is most effective when it does not simplify but rather presents learners with a desirable level of difficulty.