Driving anger seriously compromises road safety. Unfortunately, the majority of studies on the subject have been concerned mainly with drivers in the U.S. and some European countries. Although many media reports have suggested that its severity is increasing, the problem of road rage in many Eastern countries including China has received little research attention so far. This thesis, comprising five studies, investigates the determinants and consequences of driving anger using a sample of Chinese drivers. In the first study, a section of the literature on the effects of driving anger on three types of aberrant driving behaviours (aggressive driving, risky driving, and driving errors) and on two accident-related conditions (near misses and accidents) was integrated using the meta-analysis technique. The analysis showed that driving anger significantly predicts all the three types of aberrant driving. The effects of driving anger on accident-related conditions were also found to be statistically significant. Tests on the effects of the moderating variables (age, study publication year, and participants’ country of origin) suggested that driving anger has a more detrimental impact on young drivers than on older ones when all were similarly anger-provoked. Also, the association between anger and aggression was found to decrease over time and vary across countries. The second study adapted the short Driving Anger Scale (DAS) for use with Chinese drivers and investigated the relationship between driving anger and aggressive driving. The data were collected via an Internet-based survey conducted on a sample of Chinese drivers. A three-factor DAS structure provided a good fit with data ─ the three subscales used were hostile gesture, safety-blocking and arrival-blocking. Hostile gesture anger and arrival-blocking anger were found to be positive predictors of aggressive driving whereas safety-blocking anger was a negative predictor. The results showed that, in China, the overall driving anger was lower than that in Western countries but its association with aggressive driving was stronger. The third study went further to investigate how driving anger and aberrant driving behaviours are related to accident risk by developing and testing a mediated model, in which the effects of driving anger on road accident risk are mediated by the aberrant driving behaviours. To test the validity of the model, an Internet-based questionnaire ─ which included various measures of driving anger, aberrant driving and road accident history ─ was completed by a sample of Chinese drivers. The results showed that the model fitted the data very well and indicated that the aberrant driving behaviours significantly moderates the effects of driving anger on road accident risk. The fourth study investigated how the emotional responses of drivers, from both dimensional and discrete perspectives, could be predicted by using the appraisal components of goal relevance, blame party, and certainty. Traffic scenarios representing a combination of the three appraisal components were designed and presented to 50 survey participants. The emotional responses to each scenario were measured on an Arousal-Valence emotional space and were assigned with discrete emotion labels by applying a cluster analysis. As for the dimensional model, the results showed that valence was significantly associated with the blame party as well as the goal relevance components. Likewise, arousal could be predicted by the blame party and the certainty components. As for the discrete model, it was found that driving anger was most likely to be provoked when other drivers were responsible for the adverse driving outcome. Driving fear was most commonly experienced in situations where driver safety was threatened by the driver himself/herself or by impersonal circumstances whereas driving anxiety was an outcome of uncertain arrival-blocking events caused by the driver himself/herself or some impersonal circumstance. In the fifth study, the effects of situational (state) driving anger on driving performance and allocation of driver visual attention were studied using a driving simulator experiment. A total of 24 licensed drivers, half experienced and half novices, took part in this study. The results showed that compared with emotion-neutral drivers, drivers in an angry state tended to drive faster, maintain less headway while following a lead vehicle, and accept shorter gaps while executing left-turns. Moreover, when angry, drivers tended to adopt later and harder braking in the lane merging event, indicating a failure to respond properly to an imminent crash that fell into the peripheral areas of the road. However, responses to emergency situations arising in the central areas of the road were unaffected by situational anger. Results from eye movement data revealed that angry drivers were scanning a narrower area and applied a more heuristic processing style, both of which are likely to increase the chance of missing potential hazards in peripheral areas. Furthermore, it was found that increased experience did not prepare drivers better for the adverse influences of situational anger. The findings from our work have both theoretical and practical implications. The significance and practical use of the findings from each study will be given in details in the related chapter. Overall, from a theoretical perspective, this research demonstrates the feasibility of appraisal theory in predicting driver emotional responses. On a practical level, the findings from this study provide policy makers and researchers with a deeper understanding on the role of driving anger in the causation of road traffic accident. This should assist them in determining and targeting more effective road safety intervention strategies.