This paper investigates the lexical-semantic properties of comparative adjective phrases (CAPs) in contemporary Uzbek, analysing the morphological mechanisms through which comparativity is encoded, the semantic fields across which comparative adjectives are distributed, and the stylistic functions they perform in different registers of Uzbek discourse. As an agglutinative Turkic language, Uzbek encodes comparison through a system that is morphologically uniform but semantically and pragmatically rich: the suffix -roq and the ablative postposition -dan form the grammatical core of the comparative system, while a layered set of lexical intensifiers, equative particles, and register-sensitive construction types extends its expressive range. Through close analysis of 54 attested examples drawn from literary, scientific, journalistic, and spoken Uzbek, the study identifies five principal semantic fields of comparative adjectives, four categories of pragmatically and semantically complex usage, and three major register-conditioned variation patterns. The findings reveal that the Uzbek comparative system is typologically distinctive in several respects that bear directly on contrastive analysis with English and on the teaching and translation of comparative constructions.