The three different breast cancer subtypes (Luminal, positive, and triple negative (TNBCs) display different natural history and sensitivity to treatment, but little is known about whether residual axillary disease after neoadjuvant chemotherapy (NAC) carries a different prognostic value by BC subtype. We retrospectively evaluated the axillary involvement (0, 1 to 3 positive nodes, ≥4 positive nodes) on surgical specimens from a cohort of T1-T3NxM0 BC patients treated with NAC between 2002 and 2012. We analyzed the association between nodal involvement (ypN) binned into three classes (0; 1 to 3; 4 or more), relapse-free survival (RFS) and overall survival (OS) among the global population, and according to BC subtypes. 1197 patients were included in the analysis (luminal ( = 526, 43.9%), TNBCs ( = 376, 31.4%), -positive BCs ( = 295, 24.6%)). After a median follow-up of 110.5 months, ypN was significantly associated with RFS, but this effect was different by BC subtype ( = 0.004), and this effect was nonlinear. In the luminal subgroup, RFS was impaired in patients with 4 or more nodes involved (HR 2.8; 95% CI 1.93; 4.06, < 0.001) when compared with ypN0, while it was not in patients with 1 to 3 nodes (HR = 1.24, 95% CI = 0.86; 1.79). In patients with TNBC, both 1-3N+ and ≥4 N+ classes were associated with a decreased RFS (HR = 3.19, 95% CI = 2.05; 4.98 and HR = 4.83, 95% CI = 3.06; 7.63, respectively ypN0, < 0.001). Similar decreased prognosis were observed among patients with -positive BC (1-3N +: HR = 2.7, 95% CI = 1.64; 4.43 and ≥4 N +: HR = 2.69, 95% CI = 1.24; 5.8 respectively, = 0.003). The prognostic value of residual axillary disease should be considered differently in the 3 BC subtypes to accurately stratify patients with a high risk of recurrence after NAC who should be offered second line therapies.