ObjectiveThe methods most used for patellar height measurement were compared with the plateau–patella angle method.MethodsA cross-sectional study was conducted, in which lateral-view radiographs of the knee were evaluated using the three methods already established in the literature: Insall–Salvati (IS), Blackburne–Peel (BP) and Caton–Deschamps (CD). These were compared with the plateau–patella angle method. One hundred and ninety-six randomly selected patients were included in the sample.ResultsThe data were initially evaluated using the chi-square test. This analysis was deemed to be positive with p < 0.0001. We compared the traditional methods with the plateau–patella angle measurement, using Fisher's exact test. In comparing the IS index with the plateau–patella angle, we did not find any statistically significant differences in relation to the proportion of altered cases between the two groups. The traditional methods were compared with the plateau–patella angle with regard to the proportions of cases of high and low patella, by means of Fisher's exact test. This analysis showed that the plateau–patella angle identified fewer cases of high patella than did the IS, BP and CD methods, but more cases of low patella. In comparing pairs, we found that the IS and CD indices were capable of identifying more cases of high patella than was the plateau–patella angle. In relation to the cases of low patella, the plateau–patella angle was capable of identifying more cases than were the other three methods.ConclusionsThe plateau–patella angle found more patients with low patella than did the classical methods and showed results that diverged from those of the other indices studied.