Grouping students in terms of ability and aptitude is assumed to be advantageous, given that high-performing students may be stimulated more while low-performing students may be supported more. However, studies on ability grouping often provide mixed results. Additionally, although ability grouping is a common practice in the Philippines, it has not been empirically investigated with a nationally representative dataset. Drawing on the Philippine data in PISA 2018 which contained responses from 6952 students nested in 180 schools, we found that students in schools that practice ability grouping did not perform any better than those in schools that did not engage in ability grouping. More importantly, schools which practice ability grouping had greater academic inequalities among students. Similar academic inequalities were also more prevalent for urban and more advantaged schools while less prevalent for private schools. Taken together, these results suggest the need to attend to sources of inequalities among students in a school, and to reconsider assumptions for the supposed benefits of ability grouping.