Meta-analysis of diagnostic test accuracy studies are affected by variation in the results across individual studies (heterogeneity). We consider sources of heterogeneity and some of the ways heterogeneity may affect summary measures of accuracy study results and hence meta-analysis. We also consider how analytic methods can be used to evaluate the existence of and understand the implications of heterogeneity as well as how to appropriately adjust for heterogeneity of results. This will allow for appropriate summarization of accuracy measures and for correct inferences such as comparison between different diagnostic tests.