Multiple-choice items are one of the most commonly used tools for evaluating students' knowledge and skills. A key aspect of this type of assessment is the presence of functioning distractors, i.e., incorrect alternatives intended to be plausible for students with lower achievement. To our knowledge, no work has investigated the relationship between distractor performance and the complexity of the cognitive task required to give the correct answer. The aim of this study was to investigate this relation, employing the first three levels of Bloom's taxonomy (Knowledge, Comprehension, and Application). Specifically, it was hypothesized that items classified into a higher level of Bloom's classification would show a greater number of functioning distractors. The study involved 174 items administered to a sample of 848 undergraduate psychology students during their statistics exam. Each student received 30 items randomly selected from the 174-item pool. The bivariate results mainly supported the authors' hypothesis: the highest percentage of functioning distractors was observed among the items classified into the Application category (η2 = 0.024 and Phi = 0.25 for the dichotomized measure). When the analysis controlled for other item features, it lost statistical significance, partly because of the confounding effect of item difficulty.