This study was performed to evaluate the relation between a stable personality trait, a mood state and immune response to an examination stress. A self-reported measure of emotional stability (BFQ-ES scale) was obtained in a sample (n = 39) randomly selected from 277 cadets; this personality trait was also investigated by completing a neuroticism scale (Eysenck personality inventory) and a trait-anxiety scale (STAI). Natural killer (NK) cell activity was measured at baseline, long before the examination time and the examination day. The state-anxiety scale evaluated the response to the stressful stimulus. Taking subjects all together, the academic task did not result in significant modification over baseline in NK cell activity. Subjects were then divided into three groups based on emotional stability and state-anxiety scores: high emotional stability/low anxiety, medium, and low emotional stability/high anxiety. Examination stress induced significant increases in NK cell activity in the high emotional stability/low anxiety group, no effect in the medium group, and significant decreases in the low emotional stability/high anxiety group. The repeated-measure ANOVA revealed a significant interaction of group x period (baseline vs. examination) for both lytic units and percent cytolysis. The results did not change after introducing coffee and smoking habits as covariates. Our findings suggest that the state-anxiety acts in concert with a stable personality trait to modulate NK response in healthy subjects exposed to a psychological naturalistic stress. The relation between anxiety and poor immune control has been already described, whereas the ability of emotional stability to associate with an immunoenhancement has not yet reported. The peculiarity of our population, a very homogeneous and healthy group for life style and habits, can have highlighted the role of emotional stability, and may account for the difference with other studies.