between the two groups, apart from the finding, in the that men were more likely to improve overall than women, which seems to be implicating a vulnerability factor in the case of women. In general, the findings are that there was strong statistically significant change in all clients in the direction of improvement following attendance at a tree of life workshop, which was evident three months later. When the mild cases [score less than 7] were removed, there was still a significant improvement found. All the rates of improvement were higher than those found in the previous study, which showed improvement in only 36% of cases; that is, only 36% showed a drop in scores below 7 on the SRQ-20. The results show high rates of clear improvement for all persons attending tree of life workshops; that is, 67% overall show a drop in their SRQ-20 scores, but, when those with scores less than 10 are excluded, then the effect is not as great, but still greater than that found in the previous study – 46% as opposed to 36%. The lower rates of improvement for the more severe cases are a cause of concern. Even though all the severe cases showed significant improvement (the scores on the SRQ-20 declined significantly), this was not below the threshold (scores of less than 7) for over half the cases. This speaks to the need for a referral system for tree of life processes and the need for careful follow-up of these cases as well as a second-level care system, where there is need for the availability of mental health professionals (Reeler.2008; RAU.2009). Ideally, this should be as close to the community as possible, and probably at the level of the primary care clinic. The approach can be taught to survivors, results in the formation of small group affiliations, and can form the basis of cohesive groups around which other activities can be implemented. With the enormous displacements and political polarisation that have taken place over the past none years, it cannot be assumed that communities have maintained the cohesiveness that characterised many areas in the past, and hence it may be necessary to assume that a degree of community re-building will need to take place. The strength of approaches, such as the Tree of Life, is that they rely on people from the community itself, and creates a hub from which many other activities can develop. For example, the Amani Trust, in its programmes in Mashonaland Central in the 1990s, allied group processes to community development, and facilitated the creation of community agricultural projects that were highly successful. Any community approach will need to interface with the formal health system in the end, so that the more serious cases can be referred on to professional help. However, it seems relevant to point out that both approaches can develop in parallel: improving or even creating the capacity of the formal health system to manage the needs of survivors can develop alongside the development of community-based assistance. Experience with Zimbabwean survivors in recent years has shown considerably more resilience than might have been expected from the literature, and hence there is reason to be optimistic that low-cost, paraprofessional approaches may go a considerable way to meeting the needs of the many thousands of survivors.