Abstract JBR, a classic case of a category-specific disorder for living things reported by Warrington and Shallice (Brain 1984; 107: 829-54), was reassessed to establish whether differences in concept familiarity could account for his disorder. JBR's ability to name and define living and non-living things deteriorated with decreasing levels of familiarity, but was significantly more impaired for living things in the low familiarity range; no category-specific effect was apparent for highly familiar items. Possible confounding effects arising from the greater visual complexlty and visual similarity of living things could not account for the findings. Further investigations showed that JBR's disorder for living things could not be explained in terms of a specific loss of visual feature knowledge. Normal controls also showed a disparity between their naming of living and non-living things rated equivalently for familiarity, indicating that JBR's category disorder was not necessarily pathological in nature. It is suggested that concept familiarity remains the best predictor of performance, but that present measures, which rate only the frequency of experience, appear to be relatively insensitive to differences in the familiarity of less common living and non-living things. In future, familiarity ratings which include the quality of experience might provide a more sensitive measure.