We compare alternative methods for estimating immigrant wage and employment assimilation using unique panel data over 2001–2009 for a large, nationally-representative sample of immigrants. Previous assimilation estimates have been mainly based on cross-sectional data and have therefore suffered from a range of potential biases. We find that a fixed-effects model generates estimated employment assimilation profiles that are flatter and significantly different to those produced by cross-sectional and synthetic cohort methods. However, there are no significant differences in the wage assimilation profiles across alternative methods.