Cheating, a form of academic dishonesty, is commonly regarded as a problem in science education. This inquiry theorizes cheating not as a moral failing on the part of students or a lack of surveillance by teachers but rather as a resistance to testing. Ethnographic data from a university physical science department, analyzed with Michel Foucault's theory of governmentality, suggests testing as a technique of disciplinary power to produce normalized cases, schooled subjects of a certain type. The resistance of cheating is an assertion of agency within inequitable power relations. As such, cheating and testing are mutually constituting. This inquiry aims to trouble the notion that testing is educationally beneficial by discussing how testing may be placing students in morally compromised positions and teachers in morally complicit positions.