M.A. / The aim of this study was to compare the effect of violence on the rendering of social work services in severely strife-torn areas with its effect in moderately strife-torn areas. The social workers in the Gauteng Department of Welfare and Population Development were divided into two groups according to the area in which they render services. 2. THEORETICAL PERSPECTIVES OF VIOLENCE AND SOCIAL WORK A broad theoretical outline followed which consisted of the definition of violence, the rendering of social work services and crime. Furthermore attention was given to theories of violence and different types of violence. Then a description of violence in South Africa was given, followed by a layout of the many factors contributing to violence in South Africa. This section concluded with a description of violence in the rendering of social work services nationally and internationally. RESEARCH DESIGN The research concentrated on a comparative study which was undertaken between social workers rendering services in severely strife-torn areas and those rendering services in moderately strife-torn areas with regard to the effect violence had upon them in the rendering of services to clients. RESEARCH RESULTS Six types of hypotheses were postulated beforehand and these were then compared with information that was received from respondents from the two identified areas. The results may be summarised briefly as follows: Social workers in both strife-torn areas were prevented from visiting clients; they were prevented from going to work; their vehicles were hijacked; their vehicles were stoned often; they were abducted; they had to run away or hide; their service offices were damaged; they were late for work; their service offices had to close at times; there were times they felt that their families were in danger; they didn't want to visit clients living in certain areas; violence caused them to postpone dealing with other social problems; their relationships with their clients were strained; their clients were prevented from keeping appointments; their clients were mildly injured (no hospitalisation); their clients were seriously injured or killed; the social workers had to take leave as a result of violence; some of them possess licensed firearms (11 out of 81 social workers); they can testify to incidents where people were killed or seriously injured; they felt negative about the effect violence had on their rendering of social work services; and they were positive about the proposed ways to handle violent situations. In addition, when comparing these two areas, clients known to social workers in severely strife-torn areas can testify to more incidents where they witnessed people being killed or seriously injured, than those in moderately strife-torn areas; clients themselves were also injured more in the former than in the latter areas; and the social workers in the former areas felt that their families were in greater danger than those of their counterparts in moderately strife-torn areas. But the other variables was found not to be statistically significant. 5. RECOMENDATIONS The respondents' reaction to the proposed ways to handle violent situations led to the prioritisation of the following fourteen ways in descending order of importance: two-way radios or cellular telephones in vehicles; replacement of "G" by "T" registration of vehicles; verbal telephone codes to request assistance; in-service training in order to handle violent situations; awareness of high-risk procedures, for example the removing of a child; reporting and recording incidents of violence; retreating and getting away if possible; obtaining discreet police assistance; using conventional alarm systems; the proximity of a "stand-by" person; paired home-visits; pre-arranged interview interruptions; carrying a service firearm; and supervised office interviews. The above-mentioned ways of handling violent situations were then included in four main categories to form part of a safety strategy for this specific organisation.