Mortality rates are an important source of data for French public health observatories. These data are exhaustive, permanent, and available at various geographical levels. Nevertheless, the data are biased to some extent, namely due to the lack of a detailed description of the cause of death on the death certificate, and the causes of death are not systematically recorded when a forensic investigation has been carried out (“unknown causes”). This generates in particular an underestimation of suicide mortality in France. We studied these biases in south-eastern France for the five-year period 1995-1999, and the evolution of them between 1988-1992 and 1995-1999. In comparison with the national level statistics, deaths with an imprecise cause were more frequently recorded in south-eastern France (+30%) as was for those whose cause was unknown or non-declared (+100% in the most eastern part of the area). Thus, indicators of suicide mortality should be interpreted with caution, and compared and considered in relationship to the rate of deaths linked to an “unknown” cause. The article includes a discussion of the extent to which the transfer of data from institutes of forensic medicine to the French epidemiological centre of the medical causes of death (Inserm CepiDC), and physicians’ practices play a role.