Table 5 describes the previous health status of these survivors, which indicates the history prior to the present consultation. Table 5 here Table 6 reports the symptoms currently reported by these survivors. Table 6 here FINDINGS The overall prevalence, which reflects all cases (n=79), irrespective of diagnostic category, is very high indeed. This may be due to a sampling error, but, as far as we have been able to ascertain, the trainees are selecting patients on a random basis from the outpatient attenders. This very high rate is similar to that found in a refugee camp setting, and may reflect the extreme poverty of the district, perhaps exacerbated by the current famine. It is worth commenting here that one of the highest prevalence rates previously reported for Zimbabwe was obtained from one of the two hospitalsinvolved in the project. Patients with a history of a human rights violation are a significant proportion of the morbid group, comprising 34% of all psychological disorders, whilst the overall prevalence of victims of organised violence is 21.4% of all clinic attenders, which is again very high and in the range found in the refugee setting. The violence subsample is different in some respects from other primary care samples. There is a very high proportion of males, which is different to most primary care samples, and most of these had some military experience, some pre-dating their torture, whilst for others the torture was a precipitant in their joining the guerilla forces. It is important to stress here that very few of the sample were attending for reasons related to their history of torture, and most had no idea that they could be helped or seek compensation.