Carrington's comment on our article “Gender and the impact of gun control on homicide and suicide”, recently published in Archives of Suicide Research (1996,2,223-234), is a case of an additional statistical analysis technique on the data set. Carrington uses a statistical analysis of the regression slopes for pre- and post-regressions. We used a statistical package from Doan (1990) for time series regression, using the Cochrane-Orcutt technique to correct for serial autocorrelation. We could begin to critique Carrington's comment, for example, by noting that one would expect him to cite the test he is using and give further sources of the test. There are different tests of this sort and sometimes they yield different results. Of course, there is the ubiquitous issue of statistics in all of science, what tests to use. However, truthfully, we welcome Carrington's additional finding - both males and females benefitted from gun control by drops in suicide rates and homicide rates. This finding is in keeping with our view of the impact of environmental control for violence prevention (see special issue, Archives of Suicide Research, 1998, 4, No. 1). Thus, rather than respond further, we would like to present our own further analysis of the data in keeping with Stack's critique in Archives of Suicide Research (1998, 4) of the whole area in general. Stack argues that there is relatively little work testing sophisticated multivariate models that incorporate gun availability as one factor. The main problem according to Stack, is that we - and thus, Carrington - did not control for social economic changes in Canadian society which are co-variates of the passage of the gun control.