In face-priority automatic exposure control in digital camera systems, exposure adjustment is typically made irrespective of the face skin tone, i.e. how dark or bright the face is. As a result, a face can become over exposed when it is present against a very dark background, or become under exposed when it is present against a very bright background, depending on the face skin tone and scene content. Adapting the exposure control to the face skin tone will result in well-exposed faces in various capture scenarios, and hence better image quality.This paper presents a novel face skin tone adaptive automatic exposure control solution. Using a well-trained neural network based face skin tone predictor, the likelihoods of dark and bright face skin tones are calculated. An algorithm adjusts the bounds of the target brightness of the exposure control based on the face skin tone likelihoods and a set of configuration parameters. The face skin tone adaptive brightness bounds then guide the frame exposure adjustment. Experimental results demonstrate the outperformance of the proposed solution over conventional exposure control that does not take into account the face skin tone information.