Data on global population distribution are a strategic resource currently in high demand, in an age of new Development Agendas that call for universal inclusiveness of people. However, quality, detail, and age of census data varies significantly by country and suffers from shortcomings that propagate to derived population grids and their applications. In this work we explore the improved capabilities of recent remote sensing-derived global settlement data to detect and mitigate major discrepancies with census data. Open layers mapping built-up presence were used to revise census units deemed as ‘unpopulated’ and to harmonize population distribution along coastlines. We further developed, tested and applied automated procedures to detect and mitigate these anomalies, while minimizing changes to census geometry, preserving the regional distribution of population, and the overall counts. The two procedures employed for the detection of deficiencies in global census data obtained high rates of true positives, after verification and validation. Results also show that the targeted anomalies were significantly mitigated and are encouraging for further uses of free and open geoinformation derived from remote sensing in complementing and improving conventional sources of fundamental population statistics.