A large set of candidates for singlet fission, one of the most promising processes able to improve the efficiency of solar cells, are identified by screening a database of known molecular materials. The screening was carried out through a procedure exploiting quantum chemical calculations of excited state energies, carefully calibrated against a substantial set of experimental data. We identified ∼200 potential singlet fission molecules, the vast majority of which were not known as singlet fission materials. The molecules identified could be grouped into chemical families, enabling the design of further singlet fission materials using the hits as lead compounds for further exploration. Many of the discovered materials do not follow the current design rules used to develop singlet fission materials, illustrating at the same time the power of the screening approach and the need for developing new design principles.