Estimating school completion is crucial for monitoring SDG 4 on education, and unlike enrollmentindicators, relies on household surveys. Associated data challenges include gaps between waves, conflictingestimates, age misreporting, and delayed completion. Our Adjusted Bayesian Completion Rates (ABC)model overcomes these challenges to produce the first complete and consistent time series for SDGindicator 4.1.2, by school level and sex, for 153 countries. A latent random walk process for unobservedtrue rates is adjusted for a range of error and variance sources, with weakly informative priors. The modelappears well-calibrated and offers a meaningful improvement in predictive performance.