In three experiments, young and old adults performed an encoding task on implicational sentences and were given a subsequent recall test. Experiments 1 and 2 utilized a within-subject design in which subjects performed a pleasantness-rating or word-estimation (number of words) task followed by either free or cued recall. Although there was no age difference in free recall of sentences, the young generally recalled more than the old in response to implicational cues. The pattern was similar for pleasantness-rating sentences and word-estimation sentences, despite a large main effect of encoding task. Furthermore, the age differences were not attributable to differences in education, sex, vocabulary skill, or arousal (as measured by electrodermal responsivity). In Experiment 3, a comprehension-response task replaced the word-estimation task in a between-subjects design. The earlier findings with pleasantness rating were replicated. However, no age differences in cued recall were observed with the comprehension-response task, suggesting that the older adults benefit from explicit attention to sentence implications. Age-related differences in recall are discussed in terms of deficiencies at encoding and retrieval.