We reconsider conventional suicide measurement. First, a headcount of suicide is examined relative to some other causes of death (circulatory diseases, cancer, and motor vehicle accidents). We then construct a time-series data set of an alternative measure of suicide, the potential years of life lost (PYLL) for males and females. Suicide PYLLs average 4.57% of all male PYLLs and 2.44% of female PYLLs for 1907-2005. The comparable "count" percentages are 1.85 and 0.65, respectively. These differences are widening through time. In 2005, suicide represented 3.25% of all male deaths and 0.90% of female deaths using the count measure and, using PYLLs, 11.0% and 4.96%, respectively. The two measures produce quite different indications of suicide.