Inequality of opportunity builds upon the distinction between effort (legitimate inequality) and circumstances (illegitimate inequality). This distinction is meaningful as long as effort is not influenced by circumstances. We attempt to check the magnitude of the correlation between child effort and family background when measuring inequalities of opportunity in education by using a unique survey on secondary school education in rural Bangladesh with test scores in English and Mathematics, 7 indicators of students’ effort, a large set of circumstances and state schools versus Islamic schools. We undertake a decomposition analysis both by source and subgroups (schools). Circumstances contribute up to half of the total predicted variance in both Mathematics and English test scores, whereas effort never contributes more than 30%. However, effort remains the dominant factor for within-school variations. The ratio of the effort’s impact relative to that of circumstances drops from 3:1 to 2:1 when the correlation is put on the circumstance side for English test. All in all, these results invalidate the common practice to reduce education to a circumstance when estimating inequality of opportunity in income attainment.