We show how to use a smoothed empirical likelihood approach to conduct efficient semiparametric inference in models characterized as conditional moment equalities when data are collected by variable probability sampling. Results from a simulation experiment suggest that the smoothed empirical...