Past studies have shown that, although older adults are more likely than young adults to fail various concrete operational tasks, age differences on these tasks can be eliminated through short-term training. To test the trainability of formal, rather than concrete, operations college students (N=27) were compared with older adults (N=27). Performance on a modified version of Piaget's equilibrium-in-a-balance task divided subjects into one of three training conditions: rule plus feedback, feedback, and no feedback. A comparison of pre-test and post-test performance indicated that both young and elderly adults could use the formal operations solution rule when it was provided, but subjects' successful use of the rule did not produce positive transfer on other formal operations tasks. Simple feedback on performance produced no noticeable improvement among either young or elderly adults. Results suggest the use of formal operations is tied less to age than to experiential and health variables. (NRB)