Sixteen time series measures of crime and deaths from guns are used along with 16 measures of sales and stocks of guns in the U.S. to test for causality using Granger Causality tests. Both per capita and link-relative measures are used. The results indicate that suicide appears to result from the presence of guns and gun accidents are reduced by familiarity with guns but generally neither murder nor other crimes are affected. It appears that more causal relationships are from crimes to gun ownership than from gun ownership to crimes.