Post-marketing classical Chinese patent medicines(CPMs) have been widely used in China.However，lack of high-quality evidence-based data blocks its promoting abroad.A well-designed registry-based prospective cohort study can help better understand the causal relationship between treatment and outcome in a realistic medical environment.In this paper，we took Angong Niuhuang Pills for the treatment of acute ischemic stroke as an example to discuss the key points in clinical design，statistical considerations，and quality control in the design of registry-based cohort study from the perspective of controlling selection bias，information bias，and confounding bias.