In two studies, students at California State University, Northridge wrote papers that were checked for plagiarism using plagiarism-detection software. In the first study, half of the students in two classes were randomly selected and told by the professor that their term papers would be scanned for plagiarism using the software. Students in the remainder of each class were not informed that the software would be used. The researcher predicted that students who were explicitly warned about the use of the software would plagiarize less than students who were not, but the warning had no effect. In a second study, students wrote two papers in a series. Their knowledge about plagiarism-detection software was inversely correlated with plagiarism rates on the first paper, but no correlation was found between knowledge and plagiarism on the second paper. Instead, participants were discovered to draw repeatedly from the same sources of plagiarized material across papers. (Contains 1 table.)