In this paper, we present the first attempt at modelling the impact of economic policy uncertainty on renewable energy consumption in the USA using monthly data from 1986 to 2019. By implementing recent nonparametric (nonlinear) econometric approaches, we find that our models suffer from nonlinearity and smooth, as well as abrupt structural changes. The nonparametric unit root tests indicate non-stationarity of the model variables while the cointegration suggests the presence of nonlinear cointegration. The Granger causality analyses establish robust nonlinear causation in both directions between the policy uncertainty and renewable energy variables, with one exception: from geothermal energy to the three-component index of uncertainty. The nonparametric regressions exhibit a negative long-run association between policy uncertainty and renewable energy consumption, i.e., higher economic policy uncertainty lowers renewable energy consumptions and vice-versa. These findings have robust policy implications as they underscore the importance of governments, policymakers and concerned agents to maintain uniformity in economic policy to encourage renewable energy consumption in the USA.