A 1972 follow-up of eleventh graders surveyed by Project Talent in /960 yields results broadly similar to those reported by Sewell and Hauserfor their Wisconsinfollow-up and by Alexander, Eckland and Griffin for their EEO follow-up, suggesting that neither the geographic and educational restrictions of the Wisconsin survey nor the low response rate of the EEO systematically biases their conclusions. The Talent data also provide more detailed information on teenagers' cognitive skills and educational aspirations than the Wisconsin and EEO surveys. The Talent data show that (1) conventional academic aptitude tests predict later success less accurately than academic achievement tests administered at the same time, (2) detailed measures of educational plans beyond high school predict later behavior more acc-urately than questions that merely ask high school students whether they plan to attend college and (3) measures of how much money high school students hope or expect to earn have a very weak relationship to actual earnings.