Ten subjects with WAIS Full Scale IQ scores ranging from 119 to 47 were required to discriminate between two lines of markedly different length, exposed in random order for ten different durations, by pressing one of two keys. Estimates of inspection time gamma, the rate at which sensory input is accumulated and passed to subsequent decision processes, were calculated directly from the psychometric functions obtained gamma was found to correlate negatively with subtests contributing to Performance IQ. There was a substantial degree of consistency between estimates of gamma on two different occasions and training did not appear to affect the index, although differences between the two estimates increased as the value of the first estimate increased. Mean overall response latency did not correlate significantly with intelligence and measures obtained from retarded subjects were very like those reported in similar experimental situations utilizing normal subjects. This unexpected finding is interpreted as suggesting that, in this situation, retarded subjects have been prepared to respond on the basis of less evidence than is usually required by intellectually normal subjects.