The development of more potent selective oestrogen receptor antagonists and degraders (SERDs) that can be orally administered could help to address the limitations of current endocrine therapies. We report the primary and final analyses of the coopERA Breast Cancer study, designed to test whether giredestrant, a highly potent, non-steroidal, oral SERD, would show a stronger anti-proliferative effect than anastrozole after 2 weeks for oestrogen receptor-positive, HER2-negative, untreated early breast cancer. In this open-label, randomised, controlled, phase 2 study, postmenopausal women were eligible if they were aged 18 years or older; had clinical T stage (cT)1c to cT4a-c (≥1·5 cm within cT1c) oestrogen receptor-positive, HER2-negative, untreated early breast cancer; an Eastern Cooperative Oncology Group performance status of 0-1; and baseline Ki67 score of at least 5%. The study was conducted at 59 hospital or clinic sites in 11 countries globally. Participants were randomly assigned (1:1) to giredestrant 30 mg oral daily or anastrozole 1 mg oral daily on days 1-14 (window-of-opportunity phase) via an interactive web-based system with permuted-block randomisation with block size of four. Randomisation was stratified by cT stage, baseline Ki67 score, and progesterone receptor status. A 16-week neoadjuvant phase comprised the same regimen plus palbociclib 125 mg oral daily on days 1-21 of a 28-day cycle, for four cycles. The primary endpoint was geometric mean relative Ki67 score change from baseline to week 2 in patients with complete central Ki67 scores at baseline and week 2 (window-of-opportunity phase). Safety was assessed in all patients who received at least one dose of study drug. The study is registered with ClinicalTrials.gov (NCT04436744) and is complete. Between Sept 4, 2020, and June 22, 2021, 221 patients were enrolled and randomly assigned to the giredestrant plus palbociclib group (n=112; median age 62·0 years [IQR 57·0-68·5]) or anastrozole plus palbociclib group (n=109; median age 62·0 [57·0-67·0] years). 15 (7%) of 221 patients were Asian, three (1%) were Black or African American, 194 (88%) were White, and nine (4%) were unknown races. At data cutoff for the primary analysis (July 19, 2021), the geometric mean relative reduction of Ki67 from baseline to week 2 was -75% (95% CI -80 to -70) with giredestrant and -67% (-73 to -59) with anastrozole (p=0·043), meeting the primary endpoint. At the final analysis (data cutoff Nov 24, 2021), the most common grade 3-4 adverse events were neutropenia (29 [26%] of 112 in the giredestrant plus palbociclib group vs 29 [27%] of 109 in the anastrozole plus palbociclib group) and decreased neutrophil count (17 [15%] vs 16 [15%]). Serious adverse events occurred in five (4%) patients in the giredestrant plus palbociclib group and in two (2%) patients in the anastrozole plus palbociclib group. There were no treatment-related deaths. One patient died due to an adverse event in the giredestrant plus palbociclib group (myocardial infarction). Giredestrant offers encouraging anti-proliferative and anti-tumour activity and was well tolerated, both as a single agent and in combination with palbociclib. Results justify further investigation in ongoing trials. F Hoffmann-La Roche.