In this paper, the authors use a treatment model correcting for potential selection bias to evaluate the performance effect of Business Links assistance on small firms in England.