Changes in auditory and visual recognition were studied in subjects ranging in age from 10 to 60 years. Four-word lists were presented serially at a rate of 0·75 sec. per item and followed after 1·5 sec. by a test word for recognition. Under a separate condition subjects simply named the test word as rapidly as possible. It was argued that subtracting naming latency from decision latency yielded a relatively pure measure of comparison time in memory. Age effects were found in decision latency, but these were greatly reduced by the subtraction procedure. When presentation list and test word were both given auditorily no age differences were found in corrected decision latency, but developmental changes remained when the list and the test word were presented in different modalities. It was concluded, first, that in comparison with perceptual and response factors, memory scanning time is relatively insensitive to age differences and, second, that auditory recognition involves the use of a pre-linguistic or ‘representational’ memory system which is not sensitive to age over the sampled range.