<p>Cohabiting with a partner, rather than living in a legal marriage, has been documented as a significant risk factor for intimate partner homicide. This study used national U.S. homicide data from 1990-2005 to examine changes in risk patterns. The results showed that both male and female cohabiting victims experienced a steep decline in intimate partner homicide rates. Such was the strength of this decline that by 2005 cohabiting couples were no longer at higher risk than married couples. A similar, though weaker, trend was found in Canada between 1991 and 2006. It was hypothesised that this convergence of cohabiting and married homicide rates was due to cohabiting and married populations becoming more similar in terms of socio-demographic variables. Contrary to expectation, over time, the U.S. and Canadian cohabiting populations remained younger, poorer, less educated, and less conventional, than their married counterparts. This suggests that demographic changes were not responsible for the decline in cohabiting homicide rates. Further research is required to determine whether changes in union characteristics may have driven the decline in cohabiting homicide risk.</p> / Master of Science (MSc)