Many technologically important materials, such as nuclear fuels and Li-ion batteries contain “strongly correlated” electrons. An accurate description of these materials in density functional theory (DFT) requires an improved description of the exchange interaction. The most common improvements are the DFT+U approximation and the inclusion of exact exchange though hybrid functionals. Unfortunately, DFT+U and hybrid functionals introduce many local minima (metastable states) for the self-consistent field calculation near integer occupations of atomic orbitals. Metastable states have caused a significant number of incorrect results in the literature for actinide oxides (e.g. UO2) and other materials. We will discuss these issues and possible solutions to this problem.