The success of criminal investigations into the crimes of kidnapping, terrorism, and embezzlement can be defined as the existence of three events: the identification of at least one suspect (success 1), the accusation and putting to trial of the suspect (success 2), and a guilty verdict (success 3). The results of this study show that for each crime, different types of evidence, investigative techniques, and authorities influence the probability of the three successes. In kidnapping cases, among the most relevant significant variables are two types of evidence: technical evidence and evidence from witnesses. In terrorism cases, the results show that evidence, technical and testimonial has a greater impact than in kidnapping and embezzlement cases. Among the relevant positive variables in successes 1 and 2 are three types of evidence: (1) technical evidence, (2) evidence from raids and searches, judicial inspections and reports, (3) and artists’ impressions. For embezzlement cases, among the significant variables most relevant to the three successes are three types of evidence: official reports, statement from the entity affected, public officials of the defrauded entity, eyewitnesses and relatives of the suspect, and the defrauded entity’s participation and actions throughout the process. Impunity or punishment has been a problem widely debated both in Columbia and in Latin America for many years. This study attempted to identify the variables that explained the present capacity of the Colombian judicial system to clear up and punish the crimes of kidnapping, terrorism, and embezzlement. It is the first study in Columbia that used empirical analysis to assess the success or failure of criminal investigation. Tables, figures