Voltage unbalance is a costly and potentially damaging phenomenon. It affects both distribution network operators (DNOs) and customers. As part of a move toward greater network visibility, DNOs are increasingly motivated to monitor their networks in finer detail using an increasingly diverse range of monitoring devices. If the information from monitoring devices can be collated intelligently, the impact of unbalance and other power-quality issues can be accurately estimated throughout the network. In this paper, a new methodology is presented which utilizes distribution system state estimation (DSSE) to estimate the level, location, and impact of voltage unbalance on a real distribution network. The developed methodology is novel, pulling together and advancing existing research on DSSE and unbalance. The methodology is validated using data from a real U.K. distribution network with significant unbalance. The methodology is shown to be capable of statistically estimating the level, location, and effects of unbalance within the network, even when some areas of the network are unobservable.