When estimating the average causal effect in observationa lstudies, researchers have to tackle both confounding control and outcome modeling. This is dif?cult since usually there are a large number of confounders and the true functional form in the model is not known. Propensity score is a popular approach for dimension reduction in causal inference. We propose a new semiparametric estimation strategy using B-spline based on the propensity score. We further improve the ef?ciency of the estimator by addressing the error heteroscedasticity. We also establish the asymptotic properties of both estimators. The simulation studies show that our methods compare favorably with many competing estimators. Our method is applied to data from the Ohio Medicaid Assessment Survey (OMAS) 2012, estimating the effect of having health insurance on emergency room visit for a population with subsidized insurance plan choices under the Affordable Care Act.