Motion detection represents one of the key tasks of the visual system and has motivated a large body of research. However, present results are mainly based on the use of artificial stimuli and very little is known on the response to natural stimuli. Here, we have studied such a response t for a diurnal rodent at the earliest visual stage, the retina. For doing so, we have introduced synthetic textures that preserve properties of natural images but are easily parameterizable, in particular by modulating stimulus complexity by tuning spatio-temporal frequency content. Our results show that retinal ganglion cells responsive to motion generate tuning curves that are narrower, that is more selective, with respect to the complexity of the visual stimuli. As such, we propose that the retinal response is modulated by the stimulus’ naturalness, suggesting that adaptation to the statistics of natural images already emerges at the retina.