The heats of solution and substitution for one semiconductor in another are important quantities for an understanding of heterojunction formation. Harrison’s universal‐parameter tight‐binding method (1983) is used to obtain predictions for these quantities and for cohesive energies, defect formation energies, and solid‐state core shifts as well. The cohesive energies and solid‐state core‐shifts are compared with experiment and antistructure defect energies are compared with earlier calculations by Van Vechten. Cohesive energies are accurate to about one‐half an electron volt per bond. Agreement between experimental and theoretical cohesive energy differences is better. Predicted heats of solution and substitution depend on cohesive energy differences and therefore are likely to be realiable and are in accord with experiment in the few cases where experimental values are available. Antistructure defect formation energies agree well with Van Vechten’s earlier work. Predicted solid‐state core shifts are only accurate to a few electron volts.