We developed a microhabitat sampling methodology for investigating fish habitat use by radio telemetry. The technique is based on the concept of an area of probable occurrence (APO; a circular confidence interval around an estimated fish location) with measurements of depth, velocity, substrate, and cover features to describe the habitat in the vicinity of a fish's location. We conducted a test of precision and observer agreement of data obtained by this technique and found that the statistical precision of data varied among habitat variables and was low for features that were visually estimated. However, when data were interpreted within the context of several biologically relevant research questions, observer agreement was high indicating that the data led to consistent conclusions regarding microhabitat features within an APO.