Summary A review of recent swine crossbreeding experiments is presented and integrated with earlier work. Variation among experiments in observed heterosis for specific crosses was large for reproduction and sow productivity traits. However, standard errors of estimates were also large. It seems likely that heterosis, expressed in absolute values or in percentage units, is different for specific crosses. It is unclear whether these differences are large enough to warrant the use of unique parameter values for comparisons of alternate mating systems. In addition to average heterosis values, average direct and maternal genetic effects are given.