The relationship between the use of cannabis and the decision-making processes was explored. A computerized version of the Iowa Gambling Task (Cards Software) in its normal and reverse version was used, and the Prospect Valence Learning (PVL) model, which characterize the process of decision-making based on the parameters: Recency, Consistency, Loss aversion and Utility shape, was applied. Seventy-three cannabis consumers and a control group with 73 nonconsumers participated in the study. In the normal mode, subjects in the control group scored higher than cannabis consumers. Both groups showed consistent responses and aversion to loss. Nonconsumers showed greater influence of the gain-loss frequency, while consumers were more influenced by the magnitude of the gain-loss. The influence of immediate choices was higher among consumers who showed a quick oblivion while in the control group this process was more gradual. In the reverse mode, task performance was better among control group participants. Both groups showed consistency, loss aversion, more influenced by the magnitude of the gain-loss, and low influence of immediate elections. The results show the relationship between drug use and the decision-making processes, being consistent with the results obtained in other studies where consumers had worse results than control group. Moreover, the PVL parameters allow to adequately characterize decision-making. This confirms the relationship between drug use and decision-making by either the vulnerability prior to consumption or the neurotoxicity of drugs.