most reliably reported and provide greater detail about the circumstances of the offense. Of crimes known to police in 2000, the most recent year for which Uniform Crime Reports (UCR) data are available, firearms were involved in 66 percent of the 15,517 murders, 41 percent of the 406,842 robberies, and 18 percent of the 910,744 aggravated assaults. Data from the National Crime Victimization Survey (NCVS) for 2000 indicate about 3 percent of the 260,950 rapes or sexual assaults involved the use of a firearm, although this estimate is based on 10 or fewer sample cases (Rennison, 2001).
According to the UCR, 10,179 murders were committed with firearms in the United States in 2000, corresponding to a rate of 3.6 per 100,000.8 This count is down from a historic high in 1993 of 17,046 firearm-related murders (6.6 per 100,000). Handguns were used to commit 52 percent of all homicides, and firearms of any kind were used to commit 66 percent of all homicides in that year; 14 percent were committed with knives or other cutting implements, and 7 percent were achieved with hands, feet, or other “personal weapons.”
Trends in weapon-specific homicide rates from 1976 to 2000 are shown in Figure 3-2. Handgun homicides rose until 1993 and then fell, tracking closely the overall homicide rate, while the rates for other firearms, knives, and other weapons fell steadily and closely track each other. Thus, handgun homicides accounted for virtually all of the increase in the overall homicide rate between 1985 and 1993, the year the handgun homicide rate reached its 25-year peak of 5.4 per 100,000 (an estimated 14,005 handgun homicides).
The likely use of firearms varies dramatically from one type of homicide to another. For example, in the year 2000, about 17 percent of homicides were known to have occurred during the commission of other crimes; among these, 73 percent of robbery-related homicides were committed with a firearm, but only 9 percent of rape-related homicides were committed with a firearm.
These UCR statistics differ slightly from those presented in Table 3-3. Since the UCR collects data from police sources and the NVSS from medical examiner records, the disparity between the two systems arises because of data collection differences. Despite these differences, the systems are highly concordant in their estimates of firearm-related murder. Here we present UCR-Supplemental Homicide Report data because they provide information about offenders, weaponry, and circumstances surrounding the offense—information not found in the NVSS.