Critical realism is a philosophy of science, which has made significant contributions to epistemic debates within sociology. And yet, its contributions to ethnographic explanation have yet to be fully elaborated. Drawing on ethnographic data on the health‐seeking behavior of HIV‐infected South Africans, the paper compares and contrasts critical realism with grounded theory, extended case method and the pragmatist method of abduction. In so doing, it argues that critical realism makes a significant contribution to causal explanation in ethnographic research in three ways: 1) by linking structure to agency; 2) by accounting for the contingent, conjunctural nature of causality; and 3) by using surprising empirical findings to generate new theory. The paper develops the AART (abduction, abstraction, retroduction, testing) research schema and illustrates its strengths by employing a Bourdieusian field analysis as a model for morphogenetic explanation.