Several recent studies have employed retrospective report of the timing of events at adolescence, particularly of age at first sexual intercourse. The value of such studies rests on the assumption of reliability of the report of age on the occasion. The current study provides a test of the reliability of report of age at first intercourse over relatively short periods of time and consistency of eports of ever having engaged in intercourse. Subjects were an at-risk sample of approximately 200 adolescent boys in the Oregon Youth Study. Every year from Grade 8 through Grade 11, subjects were asked if they had ever had sexual intercourse and when was the first time. Consistency of responses as to whether they had ever had intercourse was high. Report of age at first intercourse, however, appeared very unreliable. Behavioral factors were assessed for their possible association with unreliable date reporting.