The clinical phenomenology of psychiatric disorders includes both subjective psychological experiences and objective behavioral changes. Nevertheless, the diagnostic process in psychiatry is based almost exclusively on the evaluation of the psychological symptoms as voiced by the patient with virtually no use of direct observation of behavior. The weakness of such an approach to behavioral assessment has several negative consequences for clinical practice. It is difficult to estimate to what extent the findings of biological studies are confounded or invalidated by the fact that they are generally based on correlations between accurate physiological measurements and crude behavioral ratings, sometimes of the type “better/worse” or “much/less.” Another negative consequence of psychiatry's neglect for direct observation of behavior is the difficulty of integrating animal and human data about the effects of drugs on behavior. If the clinical phenomenology of mental illnesses could be reformulated in ethological terms, the same, or similar, definitions could then be applied to the development of animal models, and analogs for specific behaviors might then become more feasible. Finally, the weakness of behavioral assessment in psychiatry has negative implications for clinical practice as well. Several studies have shown that the objective and quantitative recording of patients’ behavior may sometimes yield different results from those obtained using rating scales or structured interviews. These findings cast doubt on the validity of routine psychiatric assessments and suggest caution in basing important clinical decisions (e.g., when to discharge a patient or whether to increase drug dosage) exclusively on patients’ reports of their symptoms.