In breast cancer, nine models have been developed to predict non SN status in patients with SN metastasis. Four models are nomograms: the Memorial Sloan-Kettering Cancer Center nomogram (MSKCC nomogram), the nomogram of Degnim et al. (Mayo nomogram), the nomogram of Pal et al. (Cambridge nomogram), and the nomogram of Kohrt et al. (Stanford nomogram). Three models are scoring systems: the Tenon score, the score from the M.D. Anderson Cancer Center (MDA score), and the score of Saidi et al. Finally, two are recursive partitioning tools developed by Kohrt et al. Before being used in routine, these models have to be validated in independent populations based on discrimination and calibration. However, the main issue is their clinical utility based not only on the low false negative rate but also its potential to discriminate patients with a low risk of non SN involvement. Several institutions have tested the MSKCC nomogram, with AUC ranging from 0.58 to 0.86. It was not validated by four studies which did not recommend its use even in patients with micrometastasis.The external validation of the Tenon score confirmed its relevance with an AUC of 0.82.