This dissertation focuses on the importance of lexical frequency and phonetic context in explaining a particular case of variation. The object of study is the sound /s/ in the Spanish of Barranquilla, Colombia. This dissertation examines the role of several linguistic variables that have been somewhat overlooked, including lexical frequency and the manner of articulation of the following segment. An experiment is conducted in which 33 speakers from Barranquilla, Colombia read sentences which elicit the production of /s/ across a wide range of linguistic variables. Productions of /s/ are submitted to auditory acoustical analysis and the visual inspection of spectrograms is carried out for ambiguous cases in which the strong sibilance that characterizes the /s/ is not clear. The corpus, consisting of 2,805 tokens, is submitted to a binary logistic regression using SPSS 14.0.1 and a multiple step-wise regression using GoldVarb2001. It is found that lexical frequency is the most important factor in explaining this particular sound change. Equally important, several linguistic variables that were found to be significant in previous research turn out to be insignificant once lexical frequency is considered in the analysis. The dissertation supports phonological models that are based on language usage and concludes with the implications that this experiment has on our present understanding of sound variation and on phonological theory in general.