In 1995-96 several terrorist attacks struck Paris. After that, the French government decided to optimize the service claimed to treat psychological repercussions of attacks victims. For this reason we need to better understand the psychopathology developing after these traumatic events in order to adjust the various steps of the treatment. In December 1996, a terrorist attack occurred in a Paris subway. Medical and medico-psychological teams intervened immediately on the site to help victims. Among 115 victims, 4 persons died and 35 were seriously injured. The aim of our study was to evaluate the psychological impact among a population of terrorist attacks victims by a prospective study and to identify predictive factors of posttraumatic stress disorder (PTSD). We evaluated PTSD rates at 6 and 18 months, the relationship between coping style and PTSD, and whether PTSD increased health care utilization. Two follow up evaluations were performed in the 6th and 7th month respectively, by means of self-questionnaires sent by mail. Among 115 victims of the bombing attack occurred in December 1996, the 111 survivors were asked to participate to the study. The subjects who accepted and could use French questionnaires were considered eligible for the inclusion: the main criteria of the Watson's PTSD Inventory for the specific post-traumatic symptoms were used; the Goldberg's General Health Questionnaire was used to measure the general psychopathology; to identify coping styles we used the questionnaire "Ways of Coping Check List" of Vitaliano at 6 months and the "Coping Inventory for Stressful Situations (CISS)" by Endler at 18 months; a small questionnaire was proposed to evaluate injuries, hospitalization and specific treatment immediately or after the event. Among 70 subjects who accepted to participate, 56 (33 females) could be evaluated at 6 months and 32 (14 females) subjects at 18th months. The mean age at 6 months was 38.4 years: 41% of participants met PTSD criteria at 6 months, 34.4% still had PTSD at 18 months; at 18th months, 50% subjects presented GHQ scores higher than 3, corresponding to the cut-off revealing mental suffering; people hospitalized more than two days immediately after the event presented significantly lower PTSD scores at 6 months. General characteristics of risks factors for PTSD were: at 6 and 18 months, women and younger people presented significantly higher PTSD scores; for the CISS at 18 months Emotion-oriented coping correlated significantly with PTSD (r = 0.49, p = 0.007), while task-oriented coping and PTSD correlated negatively (r = -0.39; p = 0.04). Avoidance coping's styles were not correlated with PTSD. About predictive factors: the GHQ-12 and PTSD-I scores at 6 months were significantly correlated with PTSD-I scores at 18 months (respectively r = 0.73, p = 0.018 and r = 0.75, p = 0.0029); by a multiple regression we observed that PTSD-I score at 6 months predicted the PTSD-I score at 18 months, adjusted on sex and age. The others characteristics at 18 months were: medication use increased significantly at 18 months; people who lived another traumatic event since December 1996 presented a 18 month PTSD score higher than the other victims. In spite of the small size of this sample, the principal interest of this study is the prospective data in a population exposed by the same traumatic event. We note the high score of PTSD at 18 months. Terrorism exposure resulted in persisting PTSD in a significant proportion of victims; this was related to coping style. Moreover PTSD increased health care utilization. We discuss these results comparing with other similar populations in France, Israel and USA. We discuss overall the role of coping styles during the time after an attack; we insist on considering this aspect in the therapeutic strategies. These data contribute to inform that people with a high PTSD score at 6 months presented a high risk to suffer PTSD at 18 months. These results underline the importance of early diagnosis to propose early medical and psychological help to the victims.