The explosion at the AZF factory in Toulouse occurred on September 21, 2001. Workers and rescue personnel were particularly affected due to the large number of businesses that were partially or completely destroyed by the explosion. Following a cross-sectional survey of workers in the Toulouse metropolitan area (Year T0), a cohort of volunteers was established in May 2003 with the objective of describing the health and socio-occupational consequences of the explosion over a five-year period. After the enrollment phase, which took place between May 2003 and July 2004 (Year T1), follow-up of cohort participants was conducted via a self-administered questionnaire mailed to each volunteer on the anniversary of their enrollment in the cohort. In total, 1,047 men and 1,209 women responded to the self-administered questionnaire exploring career trajectory, mental health, and sensory disorders during the follow-up year from May 2004 to April 2005 (Year T2). Assessment of depression using the CES-D showed a prevalence of 34% among men and 50% among women. These prevalences vary by socio-occupational category: they are higher among manual workers (48% among men and 55% among women) and artisans and merchants (43% and 53%) than among managers (22% and 42%). The symptoms of post-traumatic stress (PTS) assessed by the IES-R show prevalences of around 15% among men and 22% among women. These PTSD prevalence rates also vary by socio-professional category: the categories of manual workers (26% and 37%) and artisans and shopkeepers (27% and 44%) are more affected than managers (6% and 12%). These prevalence levels of mental health disorders, four years after the disaster, demonstrate the significance of these disorders; they are consistent with previous results published in the "AZF" health cohort. Regarding reported hearing disorders, our results highlight the prevalence of tinnitus (33% among men and 27% among women) and hyperacusis (30% and 39%). These high proportions are likely due to the data collection method used during the May 2004–April 2005 follow-up year (self-administered questionnaire), which differs from that used at enrollment (administered questionnaire). Analysis of the upcoming years of follow-up using self-administered questionnaires will allow for the best possible comparison and monitoring of the progression of hearing disorders. The "AZF" health cohort is scheduled for a five-year follow-up period. The results from the May 2004–April 2005 follow-up year (T2) corroborate those from baseline (T1), as they demonstrate the persistence of these disorders four years after the disaster. These results are part of a longitudinal follow-up study that will assess the medium-term consequences of the explosion, primarily in the areas of mental health and hearing disorders. (R.A.)