Studies on validation of nursing diagnoses have been strongly influenced by the methodological framework developed in the 1980s. The suitability of these models to validate elements that make a nursing diagnosis has undergone several criticisms. This article aims to discuss these limitations and describe alternative approaches to overcome the criticism of traditional models. A critical analysis based on recent published literature on new approaches in research diagnoses was taken as basis for the discussions presented. Alternative methods are described for the concept analysis applied the procedures for validation of nursing diagnoses; methods for diagnostic content analysis by proficient judges, and methods for clinical validation of indicators and etiological factors. Conclude that there are few studies with these new approaches and the complexity of the analysis is an aspect that hinders its implementation.