BACKGROUND: Disparity in utilization of reproductive healthcare services between the urban poor and the urban non-poor households in the developing nations is well known. However, disparity may also exist within urban poor households. Our objective was to document the extent of disparity in reproductive healthcare utilization among the urban poor and to identify the socio-demographic determinants of underutilization with a view to characterizing this vulnerable subpopulation. METHODS: A survey of 16,221 households was conducted in 39 clusters from two large urban poor settlements in Delhi. From 13,451 consenting households, socio-demographic data and information on births, maternal and child deaths within the previous year was collected. Details of antenatal care (ANC) was collected from 597 pregnant women. Information on ANC and postnatal care was also obtained from 596 recently delivered (within six months) mothers. All data were captured electronically using a customized and validated smart phone application. Households were categorized into quintiles of socio-economic position (SEP) based on dwelling characteristics and possession of durable assets using principal component analysis. Potential socio-demographic determinants of reproductive healthcare utilization were examined using random effects logistic regression. RESULTS: The prevalence of facility based birthing was 77% (n = 596 mothers). Of the 596 recently delivered mothers only 70% had an ANC registration card, 46.3% had ANC in their first trimester, 46% had visited a facility within 4 weeks post-delivery and 27% were using modern contraceptive methods. Low socio-economic position was the most important predictor of underutilization with a clear gradient across SEP quintiles. Compared to the poorest, the least poor women were more likely to be registered for ANC (OR 1.96, 95%CI 0.95-4.15) and more likely to have made ≥ 4 ANC visits (OR 5.86, 95%CI 2.82-12.19). They were more likely to have given birth in a facility (OR 4.87, 95%CI 2.12-11.16), to have visited a hospital within one month of childbirth (OR 3.18, 95%CI 1.62-6.26). In general, government funded health insurance and conditional cash transfers schemes were underutilized in this community. CONCLUSION: The poorest segment of the urban poor population utilizes reproductive healthcare facilities the least. Strategies to improve access and utilization of healthcare services among the poorest of the poor may be necessary to achieve universal health coverage.