The network scale-up method enables researchers to estimate the size of hidden populations, such as drug injectors and sex workers, using sampled social network data. It offers advantages over other size estimation techniques, but the basic scale-up estimator depends on problematic modeling assumptions. We propose a new generalized scale-up estimator that does not suffer from these problems. The new estimator can be used in realistic settings, like populations with non-random social mixing and imperfect social awareness, and it can accommodate data collection with complex sample designs and incomplete sampling frames. However, the generalized scale-up estimator also requires data from two samples: one from the frame population and one from the hidden population. In some situations these data from the hidden population can be collected by adding a small number of questions to already-planned studies. For other situations, we develop interpretable adjustment factors that can be applied to the basic scale-up estimator. We conclude with practical recommendations for the design and analysis of future studies.