In Liberia, 75% of those who died from 2014 Ebola epidemic were women and the effects of this gruelling epidemic were more severely felt by pregnant women. This immediately raised fears about the long-term impacts of the epidemic on maternal and child health. As part of a larger study, this paper uses Andersen’s behavioural model of health care utilization and Goffman’s stigma theory to explain the timing and utilization of maternal health services before the outbreak of the Ebola epidemic as a background to the potential long-term effects on maternal health. We conducted survival and multiple regression analysis using the 2007 (N = 3524) and 2013 (N = 5127) Liberia’s Demographic and Health Survey (LDHS) data. Our sample consisted of women of reproductive age (15–49 years) that had given birth in the last five years preceding the survey year. The findings show that from 2007 to 2013, there was an overall improvement in the timing of first antenatal care (ANC) visits (TR = 0.92, p < 0.001), number of ANC visits and delivery with skilled birth attendants. The results also show county and regional disparities in the utilization of ANC services with South Eastern A region emerging as a relatively vulnerable place. Also, access to ANC services defined by distance to a health facility strongly predicted utilization. We argue that the Ebola epidemic likely eroded many of the previous gains in maternal health care, and may have left a lingering negative effect on the access and utilization of maternal health services in the long-term. The study makes relevant policy recommendations.