Abstract. The aim of this study was to evaluate the influence of Stress Inoculation Training (SIT) on the anxiety level measured by State-Trait Anxiety Inventory (STAI) both as temporary/emotional state anxiety (X1) and stable personality trait anxiety (X2) in soldiers preparing for their mission in Afghanistan (ISAF).

Method. 118 soldiers were randomly selected from the contingent that consisted of 1500 soldiers and split into two, equinumerable groups - experimental (E) and control (C). Both groups listened to a lecture on the nature of stress, its symptoms and coping with stress. They also filled in the following inventories: STAI, PCL-M, BDI-2, CISS, NEO-PI-R, FCB-TI and TAS. Soldiers from the E group - split into four subgroups - took part during the next 5 days in 10 SIT sessions according to the methodology of Virtual Reality Medical Centre, San Diego. At the same time soldiers from the C group took part in their scheduled training in their military area. After completing the SIT both groups filled in STAI. After the end of their deployment in Afghanistan the soldiers filled in STAI, PCL-M and took part in a structured interview.

Results. The statistical analysis of STAI results shows that: 1) Before SIT, there were no statistically significant differences in STAI and other tests results in both E and C groups. 2) After SIT, both X1 and X2 values in E group were significantly lower (p = 0.04). 3) In the C group which did not take part in SIT, there were no statistically significant differences in X1 value after 5 days; however, there was a statistically significant decrease in X2 value (p = 0.01). 4) After deployment, both X1 and X2 values in the E and C group were significantly lower comparing to X1 and X2 values before SIT (p < 0,05) 5) After deployment, X2 values in the E and C group were significantly lower comparing X2 values measured after SIT (p < 0.01) 6) There were no statistically significant differences in X1 values after deployment compared to values measured after SIT in the E and C group 7) There were no statistically significant differences in X1 and X2 values between the E and C group before, after the experiment and after deployment.

Conclusions. Given the equivocal results of the experiment there is a need for a further study or a deeper analysis.