Using a cross-lagged analytic framework, the present study examined (1) the relative salience of prior academic personal bests (PBs) in predicting subsequent engagement and achievement compared with (2) the relative salience of prior engagement and achievement in predicting subsequent PBs. Academic PBs, engagement, and achievement measures were administered to 1866 high school students at two time waves across a one-year interval. Path models suggest the salience of prior academic PBs over subsequent engagement and achievement, and in some instances, evidence of reciprocal effects. The findings hold substantive, applied, and methodological implications for researchers and practitioners seeking to improve students' academic development through academic PBs. (Contains 2 tables and 2 figures.)