The joint-rater and test-retest reliability study of two translated versions of the SADS-LA (Schedule for Affective Disorders and Schizophrenia--Lifetime version--modified for the study of anxiety disorders), one in French and the other in German, have been tested in family study settings, in a sample of patients and first-degree relatives. The test-retest reliability study demonstrated that identification of major affective disorders and schizophrenia was performed with sufficient reliability; however, diagnoses of subtypes of major disorders (e.g. bipolar II disorder) and identification of minor disorders was less reliable. The implications of these findings in phenotype identification during family studies in psychiatry are discussed.