Adherence to rationally prescribed medications is essential for effective pharmacotherapy. However, widely variable adherence to protocol-specified dosing regimens is prevalent among participants in ambulatory drug trials, mostly manifested in the form of underdosing. Drug actions are inherently dose and time dependent, and as a result, variable underdosing diminishes the actions of trial medications by various degrees. The ensuing combination of increased variability and decreased magnitude of trial drug actions reduces statistical power to discern between-group differences in drug actions. Variable underdosing has many adverse consequences, some of which can be mitigated by the combination of reliable measurements of ambulatory patients' adherence to trial and nontrial medications, measurement-guided management of adherence, statistically and pharmacometrically sound analyses, and modifications in trial design. Although nonadherence is prevalent across all therapeutic areas in which the patients are responsible for treatment administration, the significance of the adverse consequences depends on the characteristics of both the disease and the medications.