The memory segment of the semiconductor industry experienced two times of technological leadership change, from the US to Japan in 1982 and from Japan to Korea in 1993, and since then there has been no leadership change for more than two decades. This paper examines reasons why this happened and discusses implications for catch-up strategies and changing windows of opportunities for latecomers. It identifies the catch-up strategy imperative in the memory industry as ‘dynamic catch-up strategy’ and investigates how it worked for Japan and Korea. It then explores reasons why the strategy later became more difficult to implement successfully.