The authors investigated ways of encouraging students to consider more counterarguments when writing argumentative texts. One hundred eighty-four undergraduates wrote essays on TV violence. In Experiment 1, students given specific goals generated more counterarguments and rebuttals than controls. In Experiment 2, some participants were provided with a text outlining arguments/counterarguments; some were also asked to write a persuasive letter. Prior attitudes toward the topic were also measured. Persuasion instructions negatively affected and text (without persuasion instructions) positively affected counterargumentation and the overall quality of arguments. Text was only effective, however, for students with less extreme prior attitudes. The danger of using persuasion goals and the advantages of using more specific goals (with text) are discussed.