The present research compared the relative effectiveness of an attribution strategy with a persuasion strategy in changing behavior. Study 1 attempted to teach fifth graders not to litter and to clean up after others. An attribution group was repeatedly told that they were neat and tidy people, a persuasion group was repeatedly told that they should be neat and tidy, and a control group received no treatment. Attribution proved considerably more effective in modifying behavior. Study 2 tried to discover whether similar effects would hold for a more central aspect of school performance, math achievement and self-esteem, and whether an attribution of ability would be as effective as an attribution of motivation. Repeatedly attributing to second graders either the ability or the motivation to do well in math proved more effective than comparable persuasion or no-treatment control groups, although a group receiving straight reinforcement for math problem-solving behavior also did well. It is suggested that persuasion often suffers because it involves a negative attribution (a person should be what he is not), while attribution generally gains because it disguises persuasive intent.