Data exclusivity is designed to protect undisclosed clinical data from disclosure and unfair commercial use for a certain exclusive peri-od of time. Generated by innovators, such test data are submitted to the health authority for the examination of the safety and effectiveness of a new drug before marketing. Data exclusivity is granted to the innovators for their considerable efforts put forth in generating those test data. Generally, no one can rely on those submitted data for a subsequent application for a new drug approval. Such new type of intellectual property firstly appeared in Article 1711 of the North American Free Trade Agreement, and was incorporated under Article 39.3 of the TRIPS Agreement and subsequent bilateral free trade agreements. However, the concept of “unfair commercial use” has never been clearly defined un-der the TRIPS Agreement, and the meaning of “non-reliance” has been interpreted differently from country to country. As a leading case on data exclusivity, the Canadian case, Bayer v. Canada, offered useful resources to clarify the issues surrounding the meaning of “non-reliance”. This study attempts to explore the concept of “non-reliance” under the pharmaceutical data exclusivity regime based on Bayer v. Canada. This study suggests a possible compromised interpretation of “non-reliance” with a view to balance pharmaceutical innovation and public health, and argues that Article 40-2 of the Pharmaceutical Affairs Act is consistent with the above compromised interpretation. This study also proposes several options for amendments.