Abstract
A semiempirical prediction of the occurrence of resonances in heavy-ion reactions, based on an estimate of the spreading width of the resonances via level-density calculations, is attempted. The calculations are applied to heavy-ion reactions leading to composite nuclei in the s-d-shell region (24<or=A<or=32). The results agree well with existing experimental data and yield predictions on hitherto unobserved regions of resonances.