A strong relationship exists between household gun ownership rates in the US since 1970 (GSS, downloaded April 2014, variable "owngun") and homicide rates (UCR, "Murder and nonnegligent manslaughter rate").
A dominant argument from groups advocating for a strong libertarian gun public policy approach (relatively few limitations, if any, on gun rights as determined by the 2010 Supreme Court decision in McDonald v Chicago) is that more guns make for safer communities. However, a macro-level analysis of national trends appears to show that as household gun ownership rates drop, that homicide rates also drop (all other crimes have also been dropping since around 1990, as reported by the FBI's Uniform Crime Report). While gun sales have allegedly gone up (as tracked by background checks, since gun sales cannot be officially tracked), as have gun manufacturer profits, research indicates that fewer households are owning guns, but households that own guns have more of them.
Correlation doesn't imply causation, but this relationship is strong (r = -0.89; a negative correlation below -0.7 would be considered very strong, and this is far beyond that threshold). Similarly, the chart is a nice visual depiction of the relationship. The data plotted for gun ownership represents the response rate to the question "Do you happen to have in your home any guns or revolvers" for those who answered "NO." Therefore the "increasing" values of blue line represents households that do not own a gun.
GSS data was collected approximately every other year on the question of gun ownership between 1972-2012 (23 of 40 years). Intermediate estimates are plotted based on polynomial regression. Data available on request.