I n northern Mexico, the Saltillo pinto bean variety is one of the preferred varieties since the tolerance to accelerated browning of the grain provides farmers with a wider margin for the negotiation and sale of the product. The commercial success of the Saltillo pinto and its long shelf life allow farmers to obtain greater economic benefits derived from bean production (SánchezValdez and Collaborators; 2009). The Durango Institute of Agricultural and Livestock Forestry Research (INIFAP) registered the variety pinto Centauro that was generated through a genetic cross from pinto mestizo and pinto Saltillo; this variety has registered average yields of 1220 kg / ha. When comparing the pinto Centauro with the pinto Saltillo, it was found that Centauro was earlier (90 to 94 days) and presented larger grain size, the average height of the canopy is 32 cm and the guide grows up to 68 cm (Rosales-Serna and Collaborators; 2012). Bean varieties (Phaseolus vulgaris) in the last 30 years have shown a decrease in their average yields per hectare, generating economic damage to producers in the social sector. The objective of this study is to present an alternative to reduce the variability and / or eliminate the failures / defects throughout the process, through the analysis of statistical data of the product, the bean varieties in their variables plant height, number of green beans and dry grain production, using the minitab statistical package and various quality tools that will allow us to identify, standardize and control the process, starting from the DMAIC stages. The experiment was developed in the summer of 2018 in the experimental field San Antonio de los Bravos of the UAAANUL in a geographic location of North Latitude: 250 33´ 21 ”, West Longitude: 1030 22´ 36. With three varieties of beans (Phaseolusvulgaris): Pinto Centauro (A), Pinto Saltillo (B) and Pinto Criollo (C) by means of a random block design with three repetitions each. The three varieties of beans analyzed in the research (pinto Saltillo, pinto Centauro and pinto Criollo) The three functions created presented a correlation coefficient of 0.98, which is explained in a very good correlation between the time of the vegetative cycle and the height of the plant. In relation to the variables: number of pods, dry grain production per block and average dry grain production per plant, there was no statistical difference between the varieties.