Abstract Background Most previous studies on healthcare service utilization in low-income countries have not used a multilevel study design to address the importance of community-level women’s autonomy. We assessed whether women’s autonomy, measured at both individual and community levels, is associated with maternal healthcare service utilization in Ethiopia. Methods We analyzed data from the 2005 and 2011 Ethiopia Demographic and Health Surveys (N = 6058 and 7043, respectively) for measuring women’s decision-making power and permissive gender norms associated with wife beating. We used Spearman’s correlation and the chi-squared test for bivariate analyses and constructed generalized estimating equation logistic regression models to analyze the associations between women’s autonomy indicators and maternal healthcare service utilization with control for other socioeconomic characteristics. Results Our multivariate analysis showed that women living in communities with a higher percentage of opposing attitudes toward wife beating were more likely to use all three types of maternal healthcare services in 2011 (adjusted odds ratios = 1.21, 1.23, and 1.18 for four or more antenatal care visits, health facility delivery, and postnatal care visits, respectively). In 2005, the adjusted odds ratios were 1.16 and 1.17 for four or more antenatal care visits and health facility delivery, respectively. In 2011, the percentage of women in the community with high decision-making power was positively associated with the likelihood of four or more antenatal care visits (adjusted odds ratio = 1.14). The association of individual-level autonomy on maternal healthcare service utilization was less profound after we controlled for other individual-level and community-level characteristics. Conclusions Our study shows that women’s autonomy was positively associated with maternal healthcare service utilization in Ethiopia. We suggest addressing woman empowerment in national policies and programs would be the optimal solution.