A paradigm for evaluating working memory was validated with the use of N-back tasks in a computer program to be applied within Functional Magnetic Resonance (RMF). The study hypothesis is based on the fact that a verbal 2-back difficulty level increases the success in the performance of the task in working memory, valid to be used in RMF. Through a quantitative cross-sectional study, the N-back paradigm was adapted to a PshicoPy computer program. The task consisted of responding to the stimulus or letter that is the same as the one that appeared two places behind (2 back). 4 blocks of 30 seconds each were presented, and 4 blocks of rest of 30 seconds; in total it lasts 4 minutes. To study the validity of the paradigm, an expert validation and a pilot test were carried out with the participation of 18 adult university subjects. The data was tabulated in a spreadsheet, using statistical contingency table tests; and of sensitivity and specificity. The results offered data greater than 80% sensitivity and specificity regarding the subject’s performance before the test, and the validity of the stimuli. It is concluded that the test can be used to assess working memory by means of a computer program, which can also be used in applications such as RMF.