The Tanaka-Johnston method is used worldwide to predict the diameter of canines and premolars not erupted for the convenience of not needing boards or x-rays for use. However, in recent years researchers from several countries have shown that when used in a different population for which it was designed, it can overestimate or underestimate the values. In Cuba, where the facial pattern of the population differs from the ideal for this method, it has been highly used, but there are very few studies where the reliability or accuracy of the predictions of the same is validated. Therefore, the objective of this research is to determine the applicability of the Tanaka-Johnston method for estimating the mesiodistal diameter of canines and premolars in patients aged 12-18 years. A descriptive and cross-cutting study was conducted from June 2019 to January 2020 with a population of 140 patients of both sexes between 12 and 18 years of age from Cuba. Measurements were made of the mesiodistal widths of the lower incisors, all canines, and premolars. Frequency distributions were made to the variables studied and the results were presented in statistical tables. The t-Student statistical test was used to verify significant differences. The main results obtained were that the Tanaka-Johnston method tends to overestimate the values for the female sex and underestimate them for the male, both between 0.2 and 0.3 mm, but this difference is not significant. It is concluded that the Tanaka-Johnston method can be applied in the population studied for the prediction of the mesiodistal width of unerupted canines and premolars.