Kessler and colleagues recently published 4 articles1-4 reporting lifetime and 12-month prevalence rates, age at onset, and estimates of treatment use for DSM-IV psychiatric and substance use disorders (SUDs) from the National Comorbidity Survey Replication (NCS-R) study. However, methods used in the NCS to collect SUDs have substantial limitations that were not mentioned in the Commentary5 or any of the 4 articles.