In an attempt to establish the validity of Topical Structure Analysis (TSA) as an assessment tool in student academic writing, this study applies TSA in both high- and low-rated comparison-and-contrast essays. Following Simpson’s (2000) model, the study consists of two parts. The first part quantitatively describes the physical structure of freshman college students’ high- and low-quality comparison-and-contrast essays, and the second part presents how the topical development is carried out in the said essays. Results show that although there is a remarkable parallel preference of topical progressions between the two groups of data, over 60 percent of independent clauses in the low-quality writing introduce new topics compared to less than 50 percent in high-quality writing samples. Two-proportion z-test shows that the difference is significant, p=.012 <.05. Therefore, it may be inferred that low-quality writing tends to introduce more new topics in the independent clauses than in high-quality writing.