The data were obtained from the Vancouver Urban Survey, which was conducted from October to December in 1983. Respondents were selected with a two-stage cluster sampling procedure. The initial stage involved the listing of household numbers and block locations for the 173,432 households within the city limits of Vancouver. This listing was based on the enumeration of households done for the 1981 census by Statistics Canada. Two hundred households were randomly selected from the list through internal sampling. Each of these selected households was then considered to be the first of a cluster of five households to be included in the survey sample, producing a total sample of 1,000 households. In the second stage, a household member was selected through the use of a randomly assigned selection grid. The sample procedure yielded 489 completed interviews. Data were analyzed for measures of social support, fear of crime, and socio-demographic variables. The findings indicate that supportive others are largely ineffective in reducing anxieties about personal safety and may even, under some circumstances, exacerbate anxieties about not only one's personal safety but also the safety of family and friends. These findings are inconsistent with much popular rhetoric that views the mobilization of social support as a highly effective means of addressing social problems. 2 tables, 4 notes, and 29 references