Despite almost universal primary education in Indonesia, and increasing female educational participation, gender differences remain in access to Indonesian education. This paper attempts to measure and explain these differences at primary and secondary level in Indonesia's provinces between 1980 and 1985. It examines the relationship between provincial school enrolment ratios for males and females and four factors: school availability, formal sector employment, ‘drop-out’ patterns and marriage patterns. School availability is found to be a strong predictor of enrolment levels, and stronger for females than for males. Relationships between enrolment patterns and the other three factors appear less clear cut. © 1991, Taylor & Francis Group, LLC. All rights reserved.