Based on a review of cases heard by a criminal court in the Paris region and of another series of unsolved cases (archived in the police departments of the same area), and on in-depth interviews with detectives, this paper analyzes the process of solving homicides. The discourses of professionals are compared to their practices, which reverses the hierarchy of three main factors of solving homicide cases: investigations at the crime scene, interviews with neighbors, and the search for witnesses. Those last two human factors are, in fact, the most important in finding suspects. This paper then proposes a typology of the reasons some homicides remain unsolved. In conclusion, this paper discusses the evolution of police practices and the representations of the impact that science and technology have on police work.