This article offers a critical review of past attempts and possible methods to test philosophical models of science against evidence from history of science. Drawing on methodological debates in social science, I distinguish between quantitative and qualitative approaches. I show that both have their uses in history and philosophy of science, but that many writers in this domain have misunderstood and misapplied these approaches, and especially the method of case studies. To test scientific realism, for example, quantitative methods are more effective than case studies. I suggest that greater methodological clarity would enable the project of integrated history and philosophy of science to make renewed progress.