The increasing use of missionary church records in studies of African human capital formation appears both promising and problematic. We engage with a recent article by Meier zu Selhausen and Weisdorf (2016) to show how selection biases in church record data may provoke overly optimistic accounts of European influences on Africa’s schooling revolution. Confronting their dataset – drawn from the marriage registers of the Anglican ‘Namirembe Cathedral' in Kampala – with Uganda’s 1991 census, we show that trends in literacy and numeracy of people born in Kampala lagged half a century behind those who wedded in Namirembe Cathedral. We run a regression analysis on decadal birth cohorts (1910s-1960s) showing that ethnic, gender and locational educational inequalities persisted throughout the colonial era. We argue that European influences on access to schooling, new labour market opportunities and women’s emancipation in colonial Uganda were uneven and exclusionary, while being mediated and sustained through a political coalition of the British colonial administration with the Buganda Kingdom. We call for a more sensitive treatment of African realities in the evaluation of European colonial legacies.