The present research work had the general objective of improving the diagnosis of a group of respiratory diseases in the CAP III Metropolitan Primary Care Center of Trujillo by implementing a data mining application. A pre-experimental degree investigation was carried out; using data collection tools such as record and observation cards. During the development of the software, the CRISP-DM methodology was used. There was a population of 56 patients, and taking a sample of 19 patients; giving as results after implementation, an increase in the level of morbidity of 4.96%, as well as a decrease in the average to identify the existence of a group of respiratory diseases, in 20 minutes. Also a reduction in the average cost of diagnosis in S / 327.95. Concluding in an improvement in the diagnosis of a group of respiratory diseases in the CAP III Metropolitan Primary Care Center of Trujillo.