ATM-Chk2 network is critical for genomic stability, and its deregulation may influence breast cancer pathogenesis. We investigated ATM and Chk2 protein levels in two cohorts [cohort 1 (n = 1650) and cohort 2 (n = 252)]. ATM and Chk2 mRNA expression was evaluated in the Molecular Taxonomy of Breast Cancer International Consortium cohort (n = 1950). Low nuclear ATM protein level was significantly associated with aggressive breast cancer including larger tumors, higher tumor grade, higher mitotic index, pleomorphism, tumor type, lymphovascular invasion, estrogen receptor (ER)-, PR -, AR -, triple-negative, and basal-like phenotypes (Ps < .05). Breast cancer 1, early onset negative, low XRCC1, low SMUG1, high FEN1, high MIB1, p53 mutants, low MDM2, low Bcl-2, low p21, low Bax, high CDK1, and low Chk2 were also more frequent in tumors with low nuclear ATM level (Ps < .05). Low ATM protein level was significantly associated with poor survival including in patients with ER-negative tumors who received adjuvant anthracycline or cyclophosphamide, methotrexate, and 5-fluorouracil-based adjuvant chemotherapy (Ps < .05). Low nuclear Chk2 protein was likely in ER -/PR -/AR -; HER-2 positive; breast cancer 1, early onset negative; low XRCC1; low SMUG1; low APE1; low polβ; low DNA-PKcs; low ATM; low Bcl-2; and low TOPO2A tumors (P < .05). In patients with ER + tumors who received endocrine therapy or ER-negative tumors who received chemotherapy, nuclear Chk2 levels did not significantly influence survival. In p53 mutant tumors, low ATM (P < .000001) or high Chk2 (P < .01) was associated with poor survival. When investigated together, low-ATM/high-Chk2 tumors have the worst survival (P = .0033). Our data suggest that ATM-Chk2 levels in sporadic breast cancer may have prognostic and predictive significance.