Due to certain inconsistencies between students' reports of monthly versus yearly drug use, further analyses of self-reported drug use by high school senior classes of 1976-1979 were conducted. Previous reports of drug use frequency during the past month were roughly three times larger than estimates based on usage reports during the past year, a consistent phenomenon for alcohol, marijuana and 10 other categories of illicit drugs. The results suggest four possible explanations of this phenomenon, including forgetting, "telescoping," developmental trends, or "senioritis." Self-reports of drug use may be, in many cases, systematically underreported; however, analyses of trends are likely to be valid because any biases are likely to be fairly constant from year to year. (Author/KMF)