AbstractDeveloping a high-performance donor polymer is critical for achieving efficient non-fullerene organic solar cells (OSCs). Currently, most high-efficiency OSCs are based on a donor polymer named PM6, unfortunately, whose performance is highly sensitive to its molecular weight and thus has significant batch-to-batch variations. Here we report a donor polymer (named PM1) based on a random ternary polymerization strategy that enables highly efficient non-fullerene OSCs with efficiencies reaching 17.6%. Importantly, the PM1 polymer exhibits excellent batch-to-batch reproducibility. By including 20% of a weak electron-withdrawing thiophene-thiazolothiazole (TTz) into the PM6 polymer backbone, the resulting polymer (PM1) can maintain the positive effects (such as downshifted energy level and reduced miscibility) while minimize the negative ones (including reduced temperature-dependent aggregation property). With higher performance and greater synthesis reproducibility, the PM1 polymer has the promise to become the work-horse material for the non-fullerene OSC community.