Despite World Health Organization recommendations that all women deliver with a skilled birth attendant (SBA), research continues to demonstrate large disparities in use of SBAs by socioeconomic status (SES). Yet few quantitative studies empirically examine the factors underlying these disparities, due in part to the fact that current models do not provide clear pathways-with measurable mediators-for how distal factors like SES may affect maternal health-seeking behaviors like delivering with SBAs. We propose the Disparities in Skilled Birth Attendance (DiSBA) framework for examining the determinants of use of SBAs. We posit that three proximal factors directly affect use of SBAs: perceived need, perceived accessibility of maternal health services, and perceived quality of care. Distal factors like SES affect use of SBAs indirectly through these proximal factors, and the effects can be measured. We test the assumptions of the DiSBA framework using data from the Ghana Maternal Health Survey. The analytic techniques we use include logistic regression with mediation analysis to examine the intervening effects. We find that our proxies for perceived access, perceived need, and perceived quality of care account for approximately 23% of the difference between women with no education and those with primary school education, and about 55% of the difference between women in the lowest wealth quintile and those in the middle wealth quintiles. This study suggests that proximal factors are worthy of increased attention in terms of measurement, data collection, analysis, programmatic efforts, and policy interventions, as these factors are potentially more amenable to change than the distal factors. The effects of proximal factors are also likely context specific, thus sufficient understanding in different contexts is essential to developing appropriate interventions.