Nine national surveys documenting patient underuse of prescription medications were examined to describe the variation and trends in that underuse and identify possible reasons for the substantially different rates that were reported. Underuse includes unfilled prescriptions, delayed therapy, reduced frequency, and lowered dosage. Rates of cost-related patient underuse in the studies ranged from 1.6 to 22 percent. Insurance coverage, level of wealth, age, and health status were the sociodemographic variables most strongly related to underuse. Seven additional factors in the design and administration of the surveys were identified as providing plausible explanations for the variance across surveys. The most conspicuous variation was between three government-sponsored periodic surveys and six generally one-time assessments, with the latter yielding higher rates and greater variance in underuse. Understanding the factors contributing to the variation in reported rates of underuse of medications is an important prerequisite for the design of effective prescription-drug benefit programs.