A writer's L1 is generally believed to influence their L2 English writing significantly. However, the extent to which L1 backgrounds influence Asian learners' L2 English writing has not been wholly elucidated due to the lack of data covering various learners in Asia. Therefore, this study analysed more than one-million-word essays written by 2,318 Asian students with 18 regional backgrounds and more than 14 L1 backgrounds, which were taken from the International Corpus Network of Asian Learners of English (ICNALE) Written Essays and the ICNALE Written Essays Plus, currently under construction. The analytical focus was whether learners with the same or similar L1 backgrounds were agglomerated in a single subcluster despite the difference in the other parameters, such as essay topics. Hierarchical agglomerative cluster analyses that focused on the surface layer (words), the deeper layer (lexicogrammatical features), and the latent layer (textual factor scores) of student writing revealed that the degree of L1-based output similarities may be much lower than generally believed, which requires us to reconsider the traditional view that a writer's L1 is an absolute factor in determining the aspects of their L2 use and establish a new analytical framework for discussion of Asian learners' L2 English.