Background: Type-D personality was suggested as a marker of poorer prognosis for patients of cardiovascular disease. It is defined by having a score of 10 or more on both sub-scales of the DS14 questionnaire, Social Inhibition (SI) and Negative Affect (NA). As Type-D was designed to predict risk, its temporal stability is of prime importance. Methods: Participants in the current study were 285 community volunteers, who completed the DS14, and other personality scales, at a mean interval of six years. Results: The prevalence of Type-D did not change. The component traits of Type-D showed rank order stability. Type-D caseness temporal stability was improved by using the total DS14 score. Using the scale's product as a criterion further improved temporal stability. Logistic hierarchical regression predicting Type-D classification from Time1 demonstrated that the best predictors were Time1 scores on NA and SI, with the character trait of Cooperation, and the alexithymia score adding some predictive power. Conclusions: Temporal stability of Type-D caseness may be improved by using a sum or product threshold, rather than the current rule. In any case, as the temporal stability of Type-D is limited, research is required to formulate the optimal timing for Type-D measurement for predictive purposes.