Results of psychophysical studies of specific reading disability, in both visual and auditory modalities, are reviewed. It is suggested that in order to account for conflicting results in visual psychophysics it is necessary to pay attention to the selection of participants in the research. Different recruitment and selection procedures appear to lead to different patterns of results. Specifically, studies performed with community samples with poor phonological decoding ability seem to show deficits in transient or magno functioning compared with normal readers, but studies employing samples from reading clinics or special schools without clear specification of characteristics of the dyslexics' reading patterns are unlikely to expose such differences. In auditory psychophysics, it is proposed that the oral language abilities of the participants play a key role in determining their performance. It is argued that in both modalities it is important to pay attention to homogeneity of samples.