ABSTRACTForty‐six children aged 12–16 were shown a page of meaningless text covered in random order by different plastic overlays, including seven that were various colours and one that was clear. By successive pairwise comparison each child selected the overlay that provided the greatest perceptual clarity of the text. The children with below‐average reading ability were more likely to chose a coloured overlay, and they reported more perceptual difficulty on tasks devised by Irlen (1983). In separate sessions with and without the overlay of their choice, the children read for 15 minutes and performed a visual search task. The overlay had little effect on reading initially, but after about 10 minutes the children who chose a coloured overlay read more slowly without the overlay than with it. These children reported more symptoms of visual discomfort and showed signs of tiring when they read without the overlay. The visual search performance of the children who chose a coloured overlay was initially impaired but improved to normal levels when the overlay was used. Fourteen children aged 8–16 acted as chronological or reading age‐matched controls, and undertook the reading and visual search tasks using a clear overlay which had no effect on pe