Business angels enjoy a strong reputation for being more efficient than other investors among policy makers, practitioners, and scholars. However, due to the limited availability of specific financial data, previous research has barely assessed the impact of angels on companies’ performance. This paper seeks to bridge this gap by providing evidence from a unique dataset made up of 432 angel-backed French companies which are compared to two control groups, one randomly selected and another one consisting of similar enterprises. This double comparison process enables us to purge our analysis of structural effect and to demonstrate the importance of the methodology in generating the sample. Indeed, the results we obtain significantly differ depending on the control group. Our results show that the positive influence of angels depends on the condition of the comparison. The set of BA-backed companies is more likely to exhibit superior performance when it is compared to a random sample whereas the companies’ performance is either identical or worse when it is compared to a sample composed of k-nearest neighbors. In addition, using a quantile regression technique makes it possible to differentiate the effect of business angels based on the distribution of the value of the growth rate.