Reversing the original C-Test (Raatz & Klein-Braley, 1981), the modified C-Test (mC-Test) deletes the "first" half (instead of the second half) of every second word (Boonsathorn, 1987). It is also known as the X-Test (Cleary, 1988; Prapphal, 1994, 1996). This web-based version was explored to investigate its reliability, face validity, criterion-related validity, and concordance with the Quick Placement Test (QPT). The participants were 585 undergraduate students studying varying academic areas, in a government university in northern Thailand during the 2010 academic year; all the students were volunteers. The instruments used included: (1) the Quick Placement Test (QPT), (2) the Web-Based mC-Test (WB mC-Test), and (3) a questionnaire concerning the face validity of the WB mC-Test. The findings revealed that the WB mC-Test had high reliability and high face validity. The Pearson correlation between the WB mC-Test and the QPT was significant at a medium level. Using a statistical model, it was found that the WB mC-Test could differentiate subjects into 4 levels based on the ALTE (Association of Language Testers in Europe): Level 0 Beginner to Level 3 Upper Intermediate. The WB mC-Test is supported to be a practical preliminary self-assessment test for EFL university students.