In credit scoring and classification, feature selection is used to reduce the number of variables input to a classifier. This can be done with a quadratic unconstrained binary optimization (QUBO) model, which attempts to select features that are both independent and influential. Quadratic optimization scales exponentially with the number of features, but a QUBO implementation on a quantum annealer has the potential to be faster than classical solvers. Tests were done using the German Credit Data from UC Irvine, and the results compared with those reported in the literature. In comparison with recursive feature elimination (RFE), a technique found in many software packages, QUBO Feature Selection yielded a smaller feature subset with no loss of accuracy. This opens up the possibility of using quantum annealers to programatically reduce the size of very large feature sets, especially as the size and availability of these devices increases.