摘要In an evolutionary perspective individuals are expected to vary the degree of parental love and care in relation to the fitness value that a child represents. Hence, stepparents are expected to show less solicitude than genetically related parents, and this lack of genetic relatedness has been used to explain the higher frequencies of child abuse and homicide found in stepfamilies. However, other factors .than non-genetic relatedness may cause this over-representation in stepfamilies. Here we use a 45-year data set of parental child homicides in Sweden to test two hypotheses related to the higher incidence in stepfamilies: 1) adults in different types of family differ in their general disposition to use violence, and 2) parents are more likely to kill stepchildren than genetically related children. Of the 152 perpetrators in biparental families there was an overrepresentation of perpetrators in step-families (n=27) compared with the general population. We found support for the first hypothesis in that both general and violent crime rates were higher in stepfamilies, both in the general population and among perpetrators of child homicide. However, we found no support for the second hypothesis because of the 27 perpetrators in stepfamilies the perpetrator killed a genetically related child in 13 cases, a stepchild in 13 cases and both types of children in one case. Moreover, out of the 12 families where the perpetrator lived with both stepchildren and genetic children, there was no bias towards killing stepchildren. Thus, we found no evidence for an effect of non-genetic relatedness per se.