In practice, vertical scales have been continually used to measure students' achievement progress across several grade levels and have been considered very challenging psychometric procedures. Recently, such practices have been drawing many criticisms. The major criticisms focus on dimensionality and construct equivalence of the latent trait or construct of a vertical scale. Using data from a large-scale reading comprehension assessment program with a vertical scale, this study investigated the dimensionality or the factorial structure for each grade and the equivalence of the factorial structure across grades. The empirical results from this study support the unidimensionality assumption of vertical scale across grades. The results provide empirical evidence regarding construct invariance of vertical scale across wide span of grades. (Contains 1 figure and 4 tables.)