The objective of this study was to account for acute stress disorder (ASD) and post-traumatic stress disorder (PTSD) morbidity in a self-report survey of traffic accident victims and to evaluate the relationship between ASD and PTSD in this sample, and furthermore, to find both a model of independent variables accounting for variance in ASD and PTSD symptom level. Ninety patients, treated at an emergency ward after traffic accidents, participated in this longitudinal self-report survey. ASD was assessed using the Acute Stress Disorder Scale (ASDS) and PTSD was assessed at 6-8 months follow-up using the Posttraumatic Diagnostic Scale (PDS). Twenty-five patients (28%) met the cutoff scores for ASDS. Fifteen patients (17%) fulfilled criteria for PTSD according to the PDS. ASD was only able to predict 50% of patients who later developed high levels of PTSD symptomatology. A model of three variables explained 35% of the variance in ASD symptom level. Two variables explained 40% of the variance in PTSD symptom level. In both regression models, dissatisfaction with social support was associated with a higher symptom level. The results from this study reflect already voiced problems with the ASD diagnosis. The lack of precision in predicting who will develop PTSD is pronounced in this study. The acute traumatic symptom level explains a large part of the variance in PTSD symptom level. However, other variables also seem to play an important role.