Few studies to date have considered the effect of substrate on the functional response of an organism feeding on prey of varying visibility. Intake rates of lone captive canaries, Serinus canarius L., were measured at varying seed densities on patches of either earth or short grass (<1 cm). Experiment 1, using pale seeds, found intake rates were significantly higher and search times significantly lower on earth than on grass. Two measures of crypticity (contrast in light reflectance as measured using a spectrophotometer and an experiment with humans) found pale seeds to be more visible on earth. The results from experiment 1 could be explained by this difference in crypsis. Experiment 2 used identical seeds to those in experiment 1 except they were dyed to match their backgrounds. The two measures of crypticity both found that black seeds were less visible on earth than green seeds were on grass. However, intake rates were still significantly higher on earth than grass. Seed colour preference, vegetation impeding movement, and differences in vigilance rates or seed accessibility could not explain this result. We discuss three other potentially explanatory mechanisms, the most likely of which was the greater surface area needed for scanning created by the structure of grass. Crucially, regardless of the mechanism(s) involved, many vegetated substrates share similar properties with grass (structural complexity and shiny surfaces which reflect light) and so the outcome of our findings are likely to extend to many natural situations. Conservationists wishing to encourage granivorous birds should consider enhancing food accessibility by providing uniform substrates, such as bare earth, for them to forage on. In addition, behaviour-based models should incorporate the effects of habitat into their equations of the functional response.