Background: Longitudinal indicators of mental health are rare in France, and suicidal behavior is often assessed solely on the basis of suicide mortality. Methods: The epidemiological value and methodological limitations of four medical-administrative databases—from which data on suicide attempts could be extracted—were examined in the Nord-Pas-de-Calais region (France): calls to the SAMU (emergency medical service) for suicide attempts (2009–2011), hospital emergency department visits with a diagnosis of suicide attempt (2012), medical and surgical hospitalizations for suicide attempts (2009–2011), and psychiatric care cases with a diagnosis of suicide attempt (2011). Results: One in two SAMU units, five out of 30 emergency departments, and all medical-surgical and psychiatric units provided usable data. In the latter two sources, a unique anonymous identifier allowed for statistics at the individual level; whereas in the first two sources, the data pertained only to suicidal acts. In 2011, the rate of calls for attempted suicide per 100,000 residents was 304, while the hospitalization rate with this diagnosis was 275. The highest rates were observed among men aged 20 to 49 and among women under 20 and aged 40 to 49. There is a high degree of consistency across sources regarding the average age (between 37.8 and 38.5 years) and gender (55.0% to 57.6% women). In 2011, the number of patients diagnosed with a suicide attempt and treated in psychiatric care was 2.6 times lower than the number of people hospitalized in medical-surgical units for a suicide attempt (3,563 vs. 9,327). Conclusion: The continuous data collection and the large volume of recorded information support the use of these medical-administrative databases in the development and evaluation of mental health policies. However, to improve this multi-source information system, it is necessary to increase the participation of emergency medical services (Samu) and emergency departments, as well as to improve the coding of the suicidal nature of poisonings by certain services that are clearly underreporting.