The goal of the present study was to develop and validate parallel tests of verbal and figural delayed-recognition memory with similar task demands and difficulty levels. Such tasks would allow examination of age differences and longitudinal age changes in visual recognition memory for two types of stimuli, activate divergent neural systems, and allow us to use the same procedures within the confines of functional neuroimaging as those we use in standard neuropsychological administration. The tasks introduced here include a delay between target presentation and test phase, are matched in difficulty, and yield moderate levels of performance. Individual and group differences in task performance were examined in 80 cognitively normal men and women in two older age groups: 60 to 69 and 70 to 85. Accuracy averaged 74% in both tasks, with lower performance in the oldest age group. Although accuracy was equivalent between tasks, subjects had a more liberal response bias in the figural than verbal task. Performance on the new recognition-memory tests was significantly related to Benton Visual Retention Test (BVRT; Benton [1963]. New York: The Psychological Corporation) and California Verbal Learning Test (CVLT; Delis, Kramer, Kaplan, & Ober [1987]. New York: The Psychological Corporation) performance measures. The absence of floor or ceiling effects, wide range of individual variability, and demonstrated concurrent validity of the present tasks suggest their potential utility in functional neuroimaging studies and in the early detection of cognitive decline.