The aim of this study was to investigate the strategies leading to test item responses in 60 young (20-25 years) and 60 middle-aged (35-40 years) adults, whose highest level of education had been either secondary, technical or university. Subjects were individually administered a 12 item test similar to Raven's Progressive Matrices, and were required to describe in detail the thinking underlying each solution. Pass rates did not differ according to age or educational status, but this masked differences between the groups in strategy choice. Older and technically educated subjects were more likely to adopt non-analytic strategies, depending on item type. It was suggested that the age differences may be precursors of later decline in old age, and that technical education which in Australia is vocationally specific, may rely less on abstract cognitive skills. References and tables are appended. (Author/LMO)