Recently, great progress has been achieved in the design and preparation of conjugated organic polymer photocatalysts for hydrogen generation. However, it is still challenging to develop an organic polymer photocatalyst with high photoconversion efficiency. Rational structure design of organic polymer photocatalysts holds the key point to realize high photocatalytic performance. Herein, a series of donor–π–acceptor (D–π–A) conjugated organic copolymer photocatalysts is developed using statistical copolymerization by tuning the feed molar ratio of pyrene (donor) to dibenzothiophene‐S,S‐dioxide (acceptor) units. It reveals that the photocatalytic activity of the resulting copolymers is significantly dependent on the molar ratio of donor to acceptor, which efficiently changes the polymer structure and component. When the monomer feed ratio is 25:75, the random copolymer PyBS‐3 of 10 mg with Pt cocatalyst shows a high hydrogen evolution rate of 1.05 mmol h−1 under UV/Vis light irradiation using ascorbic acid as the hole‐scavenger, and an external quantum efficiency of 29.3% at 420 nm, which represents the state‐of‐the‐art of organic polymer photocatalysts. This work demonstrates that statistical copolymerization is an efficient strategy to optimize the polymer structure for improving the photocatalytic activity of conjugated organic polymer catalysts.