This study focuses on the development of a driving risk model based on speed deviation. Speed data were collected via an on-site driving experiment on the Sanyo Expressway in Hiroshima Prefecture in Japan in 2009 to examine the impact of in-vehicle traffic warning information (IVTWI) on traffic safety. To compare the impact of IVTWI provision on traffic safety, we examined three driving scenarios, (i) without IVTWI, (ii) voice-based IVTWI, and (iii) voice and image-based IVTWI. Driving risk was defined by three levels (low, medium, and high) based on the magnitude speed deviation using an ordered response probit model. Results show that traffic safety could be improved by IVTWI provision. A high driving risk, as derived from the proposed method, could explain the occurrence of up to 38% of traffic accidents. The results of the constructed model confirm that the probability of driving risk could be reduced by IVTWI provision.