Research is examined on learning and memory from text across the adult life span. The outcomes of the extant studies are contradictory. Clarification of these confusing findings is obtained by examining variations in the learner variables of education level and performance on vocabulary tests. Problems are considered in equating the learner characteristics for cross-sectional samples of young and old adults. Meyer and Rice collected data on a sample of more than 400 adults. Four samples of 100 young and old adults were selected from this original sample in order to make different education and vocabulary comparisons. These samples were selected to parallel these learner characteristics in past prose learning studies with young and old adults. Findings from these comparisons show that age deficits in prose learning occur for older adults with average and low average verbal ability. However, the magnitude of these deficits decreases with increased verbal ability. In fact, depending on how investigators equate age groups on education and verbal ability, they can eliminate or produce aging deficits with these more highly verbal and educated adults. Further research in this area should specify the learner variables of age, education, and vocabulary (with comparably scaled tests) as well as text and task variables. (Author)