This paper models low pay transitions in Britain using a bivariate probit model with endogenous selection to address the ‘initial conditions’problem and parental variables as instruments. The exogeneity of the initial state is strongly rejected and results in considerable overstatement of theeffects of explanatory factors. The probability of being low paid dependsstrongly on low pay in the previous year. Restricting attention to those who remain employees results in an overstatement of the probability of the low paid moving up the earnings distribution, but is not found to have much effect on the estimated effects of explanatory variables.