A study was conducted to determine (1) the nature and range of self-testing procedures that college students use to monitor their reading comprehension; (2) whether their epistomological beliefs influence the nature of their comprehension monitoring procedures; and (3) whether the level of processing dictated by the test component of their reading comprehension plan influences their grades. Ninety undergraduates completed survey forms asking them to describe how they monitored their comprehension of textbooks. Reported criteria were classified as involving the retrieval of text propositions (knowledge standard) or the transformation of text propositions (comprehension/application standard). Students were then classified as having "dualistic" (perceiving knowledge as isolated facts and answers) or "relativistic" (perceiving knowledge as an organization of facts and concepts) beliefs about the nature of knowledge. Students were also classified as high or low inventives (reporting few or many emotional blocks, respectively). Results indicated that the dualists were significantly more likely to use the knowledge standard than were the relativists, and that students reporting the use of comprehension /application criteria earned significantly better grades than did those using the knowledge criteria. High inventives used more monitoring criteria than did low inventives and were more likely to use monitoring strategies that combine the two standards. Students using many monitoring criteria earned significantly better grades than did those using only one. (Copies of the survey forms are appended.) (Author/FL)