The baseline diagnostic evaluation for studies to examine the validity of proposed diagnostic entities (case identification for category validation) is an important issue in the 1980s, particularly as the categories of DSM-III come up for review and revision. Observations relevant to such diagnostic categorization from a two-year follow-up study of neurotic disorders are presented. Variables which relate to the accuracy of the baseline diagnostic assessment are discussed. Areas of current concern include: (1) inappropriate use of diagnostic categorization from large-scale community surveys as the baseline diagnostic assessment; (2) errors inherent in a single-visit assessment, particularly for character disorders; (3) pseudocomparability of diagnosis through misuse of standardized assessment instruments such as The Diagnostic Interview Schedule (DIS) or The Schedule for Affective Disorders and Schizophrenia (SADS). The need to develop improved assessment techniques for character disorders is emphasized. Also emphasized, when category validation is the goal, is the need to link the baseline diagnostic assessment to existing knowledge and concepts about the category under examination.