Relative energies for different multiplets of trimethylenemethane and methylenecyclopropane are calculated at MCSCF, CASPT2N, and DFT levels of theory. Comparison to the experimentally measured heat of formation for the (3)A(2)’ state and to the experimentally measured (3)A(2)’-(1)A(1) gap permits analysis of the relative importance of active space and basis set selection in the multiconfigurational methods. Such comparison also reveals that while DFT accurately treats triplet trimethylenemethane, there is a limitation in present DFT functionals with respect to accurately treating nondynamic correlation effects for closed-shell singlets in a molecule characterized by degenerate frontier molecular orbitals. Implications for calculations on larger systems are discussed.