The main objective of this study is to evaluate the correlation of pathological parameters related to NACT and subsequent outcomes. The secondary objective is to correlate classical parameters and survival. We analyzed a retrospective cohort of 142 female patients treated with NACT, with primary breast cancer diagnosed between January 2011 and December 2017. Slides were reviewed by two independent pathologists. Treatment-related parameters were the average percentage of tumor cellularity, size of largest axillary metastasis, and regression pattern in lymph nodes. For statistical analysis, Kaplan–Meier method was applied to estimate the survival probability of the sample and overall survival (OS) and cancer-specific survival (SS). The Gehan-Breslow test was applied to evaluate the hypothesis of no difference in survival curves for different groups. In univariate regression analysis of parameters related to the treatment effect, macroscopic pattern, median of cellularity, cellularity pooled in 3 groups, and median of largest lymph node metastasis had independent prognostic values for overall survival (OS) and cancer-specific survival (SS). Classical parameters such as nuclear and histologic grade, mitotic index, grouped ypTNM stage, and lymphovascular invasion were also correlated to survival. In multivariate regression analysis, cellularity group ≥ 40% had a higher chance of death compared to 0–5% cellularity group for both OS (Hazard Ratio: 6.59; 95% Confidence Interval = 2.30–18.9; p < 0.001; adjusted Hazard Ratio: 3.40; 95% Confidence Interval = 1.12, 10.4; p = 0.031). and SS (Hazard Ratio: 3.9; 95% Confidence Interval = 1.58–9.72; p = 0.003; adjusted Hazard Ratio: 4.21; 95% Confidence Interval = 1.69–10.5; p = 0.002). Also, macroscopic pattern correlated to survival in multivariate analysis. The ypN1 + 2 + 3 stage group was the classical parameter with strongest correlation to worse prognosis for both OS (Hazard Ratio: 10.5; 95% Confidence Interval = 2.45–44.6; p = 0.002; adjusted Hazard Ratio: 6.78; 95% Confidence Interval = 1.50–30.6; p = 0.013) and SS (Hazard Ratio: 3.56; 95% Confidence Interval = 1.51–8.38; p = 0.004; adjusted Hazard Ratio: 2.65; 95% Confidence Interval = 1.09–6.48; p = 0.032). Other classical parameters such as triple-negative molecular subtype, lymphovascular invasion and nuclear grade 3 correlated to worse survival. Our findings support the incorporation of the percentage of tumor cellularity in the pathological reports of surgical specimens as an independent prognostic factor for patients treated with NACT.