Experiment 1 examined the effects of three delay intervals (0, 24, and 48 h) on recognition memory performance of 5-, 8-, and ll-year-olds and adults using a forced-choice procedure. Results revealed significant drops in performance after 24 h for 5-year-olds and after 48 h for 8-year-olds, with no performance decrement for older age groups. However, observations made during the first experiment raised the possibility that performance in the forced-choice procedure may have been biased, since subjects could infer which item was presented previously from high confidence that the new item had never been presented. Since age differences in use of this inferential process may have exaggerated the performance differences observed with the forced-choice procedure, Experiment 2 employed a single-item yes-no testing procedure to assess age and delay effects on recognition memory. Performance increased with age and decreased with delay, but no differential performance decrement occurred across delay as a function of age. Together, results confirmed earlier studies in demonstrating age invariance in information loss across time.Memory development has been the focus of extensive investigation over the past decade. Of the two procedures typically used to assess memory performance, recall and recognition, more is known about the factors influencing the growth of recall memory (see Kail & Hagen, 1977, for a review). With regard to recognition memory, it has been shown that recognition performance does typically increase with age (Mandler & Stein, 1974;Nelson & Kosslyn, 1976;Perlmutter & Myers, 1974;Rogoff, Newcombe, & Kagan, 1974;Tversky & Teiffer, 1976), although recognition accuracy is quite high, even in preschool children (Brown & Scott, 1971;Corsini, Jacobus, & Leonard, 1969;Perlmutter & Myers, 1976). Increasing the similarity between targets and distractors has been shown to increase the size of developmental differences in recognition memory (Dirks & Neisser, 1977;Mandler & Stein, 1974;Tversky & Teiffer, 1976). In addition, degree of organization in a pictorial scene affected age differences found in recognition accuracy (Mandler & Robinson, 1978 nized pictures than they did on pictures of objects organized in a real-world scene. Adults, in contrast, did not differ in recognition accuracy for organized vs. unorganized scenes (Mandler & Ritchey, 1977). One potentially important variable is the length of delay between inspection and testing. If loss of information over time varies across age, developmental differences in recognition memory should vary as a function of delay interval. Fajnsztejn-Pollack (1973) compared recognition performance for 280 pictures in children 5-16 years old after delay intervals of 2, 5, 10, 20, and 49 weeks following initial inspection. Although recognition performance declined across the delay, no age differences were found in the rate at which items were lost from memory over time. However, FajnsztejnPollack's earliest assessment of delayed recognition performance was fully 2 weeks followin...