Introduction: Uncertainty exists regarding the best method for teaching thrust joint manipulation (TJM) to student physical therapists. The purpose of this study was to compare the effectiveness of Peyton's 4-step (P4) approach with the "see one, do one" (S1D1) approach for teaching students to perform a lumbar spine TJM task in an academic setting. Secondary objectives were to compare the effects of each instructional approach on students' attitudes and beliefs toward spinal TJM and on their motivation to learn to perform lumbar spine TJM. Review of literature: The S1D1 approach is used in the health care professions for teaching clinical tasks to students. It is unclear whether the P4 approach may better prepare students to practice TJM. Subjects: Student physical therapists. Methods: Using a factorial quasi-experimental design, an equal number of students were assigned to a P4 or S1D1 instruction group for the TJM task. Students' performance accuracy, time, and outcome performing TJM in an academic setting were measured. Paper surveys were used to collect data about students' attitudes and beliefs toward spinal TJM and their motivation to learn TJM. A generalized estimating equations approach was used for data analysis. Results: Fifty-eight students (29 per group) completed the study. There was an interaction between the instruction group and time on task performance accuracy favoring the P4 approach (P = .03). There was no interaction between the instruction group and task performance time, task performance outcome, attitudes and beliefs toward spinal TJM, or motivation to learn TJM (all P > .19). Discussion and conclusion: The P4 approach more effectively improved student accuracy when performing the TJM task in an academic setting than the S1D1 approach. However, no differences between instruction were found for performance time or outcome. Students reported a favorable perception of learning lumbar spine TJM. These preliminary results suggest that instructors may use the P4 approach to improve students' TJM procedural knowledge before task practice. However, limitations of the study may affect the internal validity and generalizability of results.