Critics have contended that non-iterative processes cannot be modeled in Optimality Theory. In response, Kaplan (2008) argues that OT should not be able to model non-iterative processes because all apparent non-iterativity is epiphenomenal, proposing a number of factors that may result in surface non-iterativity. In this paper we present data from three dialects of Crimean Tatar, demonstrating that labial harmony is non-iterative in the Central dialect. We examine morphological and prosodic characteristics of the language, demonstrating that there are no obvious ways to reanalyze this non-iterativity. As such, we contend that non-iterative harmony in Central CT, as well as harmony in several other languages is real and requires a formal analysis. We provide such an analysis using two different harmony-driving constraints, demonstrating that OT can in fact model non-iterative vowel harmony.