Laboratory studies have routinely demonstrated that testing often leads to greater learning and retention than repeated studying. In the classroom, this effect has been replicated with memory and application tasks. However, studies of classrooms involving mathematical problem solving are sparse and have had mixed results. This paper presents the results of a quasi-experimental study in an undergraduate science, technology, engineering, and mathematics course that investigated more frequent testing that incorporated aspects of mastery testing and second-chance testing. Students in the frequent testing cohort scored seven percentage points higher and earned twice the number of As and half the number of failing grades. The advantage of frequent second-chance mastery testing was found for both multiple-choice and free-response questions and remained after controlling for differences in student ability. Women and underrepresented minority students benefited from the altered testing environment to the same extent as the general population.