This paper presents a combined nonlinear and circular features extraction-based condition monitoring method for low speed slewing bearing. The proposed method employs the largest Lyapunov exponent (LLE) algorithm as a signal processing method based on vibration data. LLE is used to detect chaos existence in vibration data in discrete angular positions of the shaft. From the processed data, circular features such as mean, skewness and kurtosis are calculated and monitored. It is shown that the onset and the progressively deteriorating bearing condition can be detected more clearly in circular-domain features compared to time-domain features. The application of the method is demonstrated with laboratory run slewing bearing data.