The very high crime rate in South African cities is often argued to be the main reason for the country's small share of business ownership and self-employment. This paper revisits this hypothesis. We estimate the effect of crime on business ownership and performance using a matched dataset of census, survey and police data. In contrast to previous studies, which take into account perceived rather than actual crime and often focus on specific geographic areas, we do not find robust evidence that high crime rates have a negative impact on business ownership. Even though our estimate of the effect of crime is statistically significant and negative, it is economically small. Further, we find no evidence for a negative effect of actual crime on sales, profits and investment of informal businesses. This may imply that the low share of business ownership and self-employment in South Africa must have other reasons. These results may also suggest that crime may not be in general a serious threat for small businesses in low and middle-income countries, however, this needs further empirical research.