We examined stability and change in students' achievement goal orientations over varying tasks. Two naturalistic longitudinal studies were conducted in undergraduate courses. Students completed self-reports designed to measure their achievement goals. Achievement goals were measured four times: prior to two assignments and two exams. Four complementary analytic techniques were used to examine goal stability: differential continuity, mean-level change, individual-level change, and profile consistency. Results across both studies provide evidence for stability and change across tasks in achievement goal orientation, and comparisons made between similar and different tasks resulted in similar levels of change. Implications for theory and research in achievement goal orientation are discussed. (Contains 2 figures and 12 tables.)