Introduction: An adequate space analysis aims to predict the combined mesiodistal widths of unerupted permanent canine and premolars. Errors can lead to an application of inadequate and irreversible treatments. Objective: To assess the degree of equivalence between predictions given by three models (Moyers' Predictive Tables on percentile 75 and 50 and Tanaka-Johnston's equations) on the sum of unerupted teeth and measurements given by Cone Beam Computed Tomography, considered in this study as the "gold standard". Materials and Methods: The study sample comprised children (n=26) aged 8-13 years who visited the Department of Dentistry, University of Coimbra. Moyers' predictive tables and Tanaka-Johnston equation were applied to retrieve the space prediction. A cone-beam computed tomography was performed and the volumetric data analyzed. A concordance correlation coefficient between each model's predictions was applied. Results: On average, all 3 models tend to overestimate the cone-beam computed tomography readings and are not able to capture entirely sum of unerupted teeth variability. Moyers 50 shows a better distribution between over and under-estimation. Conclusion: This study suggests that Moyer's percentile 50 is the predictive model with the lowest percentage of absolute errors. This allow the orthodontist to choose the best predictive model in his daily practice. Palabras clave: Cone Beam Computed Tomography, orthodontics, predictive, mixed dentition analysis.