In 1980, a two-stage cross-sectional study on the prevalence of mental disorders was performed on a probability sample of 1,574 adult residents of two boroughs in Greater Athens served by a Community Mental Health Center (CMHC), After completion of the interviews, a "case" identification procedure was applied through the use of clinical criteria allocating each respondent to one of five categories ranging from "well" to definite "cases" (stage A). In stage B time 1 (1980 to 1981), two psychiatrists interviewed a sample of 360 respondents consisting of all the probable and definite cases together with randomly selected individuals from the other three mental status categories. In 1994 (time 2), a follow-up study was conducted to reinterview the sample of 360 respondents through the use of the Structured Clinical Interview for DSM-III-R (SCID). The follow-up search resulted in 182 baseline respondents being located alive, plus 38 certified as dead and a residual 140 (38.8% of the baseline sample) categorized as definitely unlocatable. We report results for the outcome of specific nosological entities over the 13-year period, Among the main findings, of the previously (1980 to 1981) identified cases, 42.8% were similarly diagnosed as cases in the follow-up study at time 2; 92.4% of the baseline stage B (1980 to 1981) noncases were also found to be noncases in 1994. "Caseness" was found to be associated with high mortality. Of the subjects interviewed at both cross-sections and diagnosed as having a psychiatric nosological entity at time 1 (1980 to 1981), 67.5% were found to be mentally healthy at time 2 (1994). Copyright (C) 1998 by W.B. Saunders Company.