This paper explores to what extent salience is a reliable predictor of second dialect acquisition (SDA). I apply a salience-based approach, according to which the adoption or rejection of a number of linguistic forms is determined by their (socio)linguistic properties, to data from a study of the accommodation of university students from Moravia living in Prague (Wilson 2010). Linguists have tested a salience approach in the analysis of language change, dialect levelling and long-term linguistic accommodation, advancing a set of criteria according to which linguistic features are considered “salient” or “non-salient”. I advance a framework for evaluating the salience of six Common Czech (CC) forms and test its effectiveness in predicting which, and to what degree, CC forms are assimilated. I argue that salience alone cannot explain the direction of accommodation or the intensity of SDA and that it is overridden by numerous external factors that are related both to the linguistic variables and to individual speakers.