The current models of alignment in dialogue focus on how interlocutors achieve mutual understanding in conversation through the collaborative effort of interlocutors (Clark & Wilkes-Gibbs, 1986). Alignment in conversation has also been proven to be enhanced through interlocutor’s convergence on linguistic and non-linguistic properties in dialogue (Pickering & Garrod, 2004). However, very little research has investigated how alignment occurs when one interlocutor is not at the same competence level as another interlocutor. In this experiment, we examined the rate of alignment between native and non-native English speakers and the balance of effort within a native and non-native interaction. Using a matching card game (invented by Krauss & Weinheimer,1964) native speakers, assigned to the Director role, had to describe abstract tangram figures to native and non-native speakers, assigned to the role of Matcher. The Matchers then had to find the image. The accuracy of scores, time taken, number of words produced and turns taken to complete this trial was recorded. The number of times a Director entrained on a Matcher’s reference proposal was also recorded. It was found that native to non-native pairs took relatively more effort to align in conversation compared to native-to-native control pairs. In addition, the native Directors used significantly more words to describe the images to non-native Matchers compared to native Matchers. Native speakers were also more likely to entrain on the reference proposals produced by a non-native speaker than a native speaker; hence they made more effort to align with a non-native speaker than a native speaker. This suggests that alignment between native to non-native interlocutors is more challenging and that the balance of effort between these pairs is unequal compared to that of a native-to-native interaction