To name an object, we need both to recognize it and to access the associated phonological form, and phonological retrieval itself may be constrained by aspects of the visual recognition process. This paper reviews evidence for such constraints, drawing on data from experimental psychology, neuropsychology, functional imaging, and computational modelling. Data on picture identification in normal observers demonstrate that the speed of name retrieval processes differs for natural objects and artifacts, due at least in part to differences in visual similarity between exemplars within these categories. Also, effects of variables on early and late stages of object identification combine in an interactive rather than an additive manner, consistent with object processing stages operating in a continuous rather than a discrete manner. Neuropsychological evidence supports this proposal, demonstrating that subtle perceptual deficits can produce naming problems, even when there is good access to associated semantic knowledge. Functional activation studies further show increased activity in visual processing areas when conditions stress object naming relative to the recognition of familiar object structures. These studies indicate that object naming is based on a series of continuous processing stages and that naming involves increased visual processing relative to recognition tasks. The data can be modelled within an interactive activation and competition framework.