We report two positron emission tomography (PET) studies of cerebral activation during picture and word matching tasks, in which we compared directly the processing of stimuli belonging to different semantic categories (animate and inanimate) in the visual (pictures) and verbal (words) modality. In the first experiment, brain activation was measured in eleven healthy adults during a same/different matching task for textures, meaningless shapes and pictures of animals and artefacts (tools). Activations for meaningless shapes when compared to visual texture discrimination were localized in the left occipital and inferior temporal cortex. Animal picture identification, either in the comparison with meaningless shapes and in the direct comparison with non-living pictures, involved primarily activation of occipital regions, namely the lingual gyrus bilaterally and the left fusiform gyrus. For artefact picture identification, in the same comparison with meaningless shape-baseline and in the direct comparison with living pictures, all activations were left hemispheric, through the dorsolateral frontal (Ba 44/6 and 45) and temporal (Ba 21, 20) cortex. In the second experiment, brain activation was measured in eight healthy adults during a same/different matching task for visually presented words referring to animals and manipulable objects (tools); the baseline was a pseudoword discrimination task. When compared with the tool condition, the animal condition activated posterior left hemispheric areas, namely the fusiform (Ba 37) and the inferior occipital gyrus (Ba 18). The right superior parietal lobule (Ba 7) and the left thalamus were also activated. The reverse comparison (tools vs animals) showed left hemispheric activations in the middle temporal gyrus (Ba 21) and precuneus (Ba 7), as well as bilateral activation in the occipital regions. These results are compatible with different brain networks subserving the identification of living and non-living entities; in particular, they indicate a crucial role of the left fusiform gyrus in the processing of animate entities and of the left middle temporal gyrus for tools, both from words and pictures. The activation of other areas, such as the dorsolateral frontal cortex, appears to be specific for the semantic access of tools only from pictures.