This study explores differences between EFL students’ performance in Internet-based and Paper-based Diagnostic Writing Tests. A recently developed e-learning app Rain Classroom is used to conduct the Internet-based diagnostic English writing test to 95 college students and an equivalent paper-based test is set for comparison. The students’ scores on these two tests are analyzed through t-test and no statistically significant difference is found between these two test modes, which implies that some new means should be worth trying for Internet-based diagnostic English writing assessment in college. However, to provide more precise suggestions, students in the study are further divided into three groups according to the disciplines to which they belong. With the aid of t-test and one-way ANOVA, it is revealed that students in the discipline of international business perform better in the Internet-based test, whereas students in the discipline of art design display a contrary tendency. Additionally, students from the discipline of urban planning demonstrate no salient performance difference between the two test modes. Therefore, some pedagogical implications are then made for the Internet-based diagnostic assessment for L2 college English writing.