PurposeThis paper examines the gender achievement gap in the Ethiopian secondary education system using tenth-grade national examination results for the period 2002/03–2013/14.Design/methodology/approachStatistical tests are employed to determine equality of the average male and female GPAs (grade point average), and equality of the proportions of female and male test-takers in the upper and lower achievement distributions. Further, a panel fixed effects model is used to evaluate factors that affect the regional average GPAs by gender and to determine whether those factors could explain observed gender achievement gaps.FindingsResults indicate statistically significant gender achievement gaps in all regional states and the capital city. The gap has been persistent with a minor declining trend and will not be closed if continued at the current trend. The authors also find that the regional average male and female test scores could be explained partially by factors such as pupil-teacher ratio, the proportion of rural students and access to secondary schools. These factors, however, cannot explain the observed gender achievement gap.Originality/valueThe study uses regional data collected from the annual educational abstracts published by the Ethiopian ministry of education. It investigates the trends of the gender achievement gaps across the regional states and overtime at the tenth-grade level. It depicts the persistence of the gap hindering girls from furthering their education.