Research over the past decade has provided overwhelming evidence that photoreception in the vertebrate eye is not confined to the rods and cones. The discovery of non-rod, non-cone ocular photoreceptors in mammals and fish arose from quite different lines of investigation. In transgenic mice entirely lacking functional rod and cone photoreceptors a range of responses to light, including the regulation of the circadian system and a pupillary light reflex, are preserved. Electrophysiological and imaging approaches were then able to characterise a coupled plexus of directly light sensitive ganglion cells. Most recently action spectroscopy has shown that a novel 'blue-light' sensitive photopigment based upon opsin/vitamin A (OP480) mediates these responses to light. Several candidate genes have emerged for OP480, with melanopsin being by far the strongest. A definitive link, however, between this gene and OP480 has still to be established. In contrast to the mammals, the discovery of inner retinal photoreceptors (IRPs) in fish started with the discovery of a new gene family (VA opsin). The teleost VA opsins form functional photopigments and are expressed in several different types of inner retinal neuron, including retinal horizontal cells. Recent studies have investigated the electrical properties of these photosensitive neurones, but their light-sensing role remains a matter of speculation. Thus the study of IRP is developing along quite separate lines. In the mammals the research is directed towards a molecular identification of the photopigment (OP480) and its cascade, whilst in fish the major effort is directed towards identifying a role for these novel photoreceptors using physiological approaches. The discovery of IRPs in the vertebrates tells us that despite 150 years of research, we still have much to learn about how the eye processes light.