Seventy-three percent of a sample of 46 Cambodian youth interviewed in 1984 and 1987 were reinterviewed in 1990 as part of a pretest for a multisite study of Cambodian refugee trauma now under way. An additional sample of convenience of 38 youth were also interviewed to determine reliability and validity of the diagnostic instruments chosen for the larger study. The DSM-III-R diagnosis of post-traumatic stress disorder (PTSD) was found to persist, but the symptoms appeared less intense over time. In contrast, the prevalence of depression dropped markedly since 1987. Subjects remained largely free of comorbid conditions. Diagnostic reliability and validity were satisfactory. The follow-up sample appeared to be functioning well despite their PTSD profiles. The findings are discussed in light of several current controversies surrounding the concept and measurement of PTSD.