We investigate how the high and low volatile states of oil prices affect the unemployment rate in the U.S. economy. We model the monthly unemployment rate data using a logistic smooth transition autoregressive process with allowing oil price uncertainty as a transition variable between high and low volatile states. Additionally, we calculate generalized impulse response functions (GIRFs) by considering the sign and size of the shocks and the initial history of the economy. The key findings are as follows: First, the estimated LSTAR model shows that the unemployment rate rises more persistently in the long run when oil price uncertainty is high but falls when oil price volatility is low. Second, the GIRF analysis shows that oil price uncertainty shocks significantly increase the unemployment rate in a highly volatile state but decrease in a low volatility state. Third, asymmetric effects of oil price uncertainty on unemployment also hold considering different signs and sizes of the oil price uncertainty shocks and the initial history of the economy. Robustness exercises confirmed the baseline results. These findings imply that employing a nonlinear framework would be appropriate while modeling oil price uncertainty and its impacts on economic activities.