People tend to underestimate how long it will take to complete tasks. We suggest that one reason people commit this planning fallacy is that they do not naturally “unpack” multifaceted tasks (e.g., writing a manuscript) into subcomponents (completing the literature review, general discussion, references section, etc.) when making predictions. We tested this interpretation by asking participants to estimate how long it would take them to complete one of several tasks: holiday shopping in Experiment 1, “getting ready” for a date in Experiment 2, formatting a document in Experiments 3 and 5, and preparing food in Experiment 4. Participants prompted to unpack the task provided longer—and, in Experiments 3–4, less biased—estimates of how long the task would take than did participants who did not. Experiment 5 showed that the debiasing influence of unpacking is moderated by task complexity: the more multifaceted the task, the greater the influence of unpacking.