In nowadays affluent societies, there are academic resilient and non-resilient students in the school population. Academic resilient students live in disadvantaged home environments, and they strive to achieve as top as they can academically (i.e. DRS – Disadvantaged Resilient Student). On the contrary, there are home-disadvantaged students who cannot perform at levels beyond what are predicted by their socio-economic status (i.e. DNS – Disadvantaged Non-resilient Student). In the Programme for International Student Assessment [PISA] 2015 Study, hosted by Organization for Economic Co-operation and Development [OECD], the share of DRS in the 15-year-old school population of the four regions of Mainland China (i.e. Beijing, Shanghai, Jiangsu, and Guangdong – collectively named as China (B-S-J-G)) ranks eighth amongst the 68 participating countries/economies. What schooling and learning characteristics can make a distinction between DRS and DNS who are studying in the relatively affluent regions in China? This study seeks to examine factors of descending order of importance discriminating between the DRS and DNS in China (B-S-J-G). The data are drawn from the PISA 2015 Study, the major assessment domain of which is scientific literacy. In accordance with the PISA’s definition of academic resilience, the participants consist of 2,450 15-year-old DRS and DNS students. This study deploys Classification and Regression Tree (CART), a versatile data-mining tool, to find out the most important factors amongst the hundreds of schooling and student learning variables collected in PISA 2015. In descending order of relative importance, the six factors determining whether a student can be accurately predicted as a DRS rather than a DNS are: Percentage of teachers in school with a Master’s degree, Environmental awareness, Science learning time per week, Student expected occupational status, Percentage of teachers in school with a Bachelor degree, and Number of learning domains with additional instruction. Based on the empirical findings, educational practitioners are informed on how to help DNS perform in science at academic levels better than what are predicted by their socio-economic status, e.g. fine-tuning of curriculum and instruction in science (with a focus on interdisciplinary scientific enquiries, stimulation of students’ environmental awareness and raising of expected occupation status in the communities they live) and upgrading of teacher qualification to the master’s level.