Purpose: The purpose of this paper is to examine the earnings differentials between the locals and the rural-urban migrants in urban labor market in Vietnam. Design/methodology/approach: Using the new Vietnam Rural-Urban Migration Survey 2013 (VRUMS2013) that is specifically designed for rural-urban migration, the author applies Appleton et al.'s (1999) procedure correcting for potential selectivity to decompose the offered earnings gap between the locals and the rural-urban migrants into within- and between-occupation pay differential. Bootstrapping is used to derive the standard errors for the decomposition results. The author further applies the propensity score matching (PSM) method to check whether the results are robust by restricting the sample to the "common support." Findings: Within-job difference, particularly, the favorable treatment toward urban workers contributes significantly to the overall and total unexplained earnings gap. Further, between-job pay differential attributed to the over-representation of urban workers in high-paying job also helps to widen the gap. These results are robust restricting to the "common support" sample using PSM. Research limitations/implications: Due to the sample size, occupations are only classified into three broad categories. Finer classification will allow a better comparison between the contributions of between and within-occupation to earning inequality. The data are only limited to a few cities and do not include other urban centers that also receive rural-urban migrants. Practical implications: Policies to promote equal pay and alleviate within-job "discrimination," especially the preferential treatment favoring the locals (rather than to provide equal access to different jobs) are crucial for migrants' labor outcome. Moreover, this study can, to some extent, be seen as a timely contribution for the debate on household registration reform in general and in Vietnam specifically. Given China's announcement to grant permanent household registration (hukou) to unregistrated migrants in late 2015, investigating whether there is a two-tier labor market in the cities in Vietnam is particularly important for the ongoing debate regarding future of household registration system (ho khau). Originality/value: This is the first study in Vietnam on rural-urban migration and occupation segregation - an area that has been relatively less well studied in developing/transitional countries. Vietnam is also one of the few developing countries who have household registration system in place. This has made it an interesting case. The author uses a new survey data to apply the Appleton et al. (1999) decomposition on the offered wage gap rather than observed wage gap. Standard errors of the decomposition results are bootstrapped and a robust check using propensity score method is conducted.