The line “knives kill three times more people than rifles” gets repeated a lot, usually as a quick comparison meant to sound decisive. To check whether it’s true, I’m going to stick to the FBI’s latest weapon breakdown in its national homicide reporting and, just as importantly, explain what the FBI means by “knives or cutting instruments” and “rifles,” because those definitions shape the comparison.
First, a quick note on what “latest” means in FBI crime data right now. The FBI’s national crime statistics have been transitioning from the long-running Summary UCR program (including the well-known “Crime in the United States” tables) to NIBRS (the National Incident-Based Reporting System). In recent years, participation and coverage have changed as agencies onboarded to NIBRS, so year-to-year comparisons can be affected by reporting coverage—not just real-world changes. That doesn’t make the weapon counts useless, but it does mean any single-year claim should be presented with some caution and preferably with context.
How the FBI defines the categories
1) “Knives or cutting instruments”
In the FBI homicide weapon tables, this category is broader than “kitchen knife” or “pocketknife.” It’s intended to capture killings committed with edged or cutting tools. It does not include every possible sharp object (because some incidents end up in other buckets), but it does represent a wide set of blade-type implements that can be used to stab or cut. This breadth tends to make the knife category relatively large compared to more narrowly defined firearm subcategories.
2) “Rifles”
The FBI’s “rifle” category is also specific. It counts homicides where the weapon was identified as a rifle. It does not include:
– Handguns
– Shotguns
– Firearms where the type is unknown
– “Other guns” depending on how the incident was coded
So “rifles” is a subset of firearm homicides, not a stand-in for “guns” overall. That’s a key reason why comparing “knives” to “rifles” can create a misleading impression if someone mentally substitutes “rifles” for “firearms.”
3) What’s missing from the simple comparison
The knife-versus-rifle soundbite ignores other large categories that can also be compared:
– “Handguns” (typically the largest identified firearm category)
– “Firearm, type not stated/unknown”
– “Personal weapons” (hands, fists, feet, etc.)
– “Blunt objects,” “poison,” and other mechanisms
If the goal is to understand homicide risk or trends, the rifle category alone is not the right yardstick for “gun violence,” even though it can be relevant for narrower debates.
Does the “three times more” claim hold up?
When you look at the FBI’s recent national homicide weapon breakdowns, knives frequently do outnumber rifles in raw homicide counts, and in many years the knife total has been roughly on the order of a few multiples of the rifle total. In other words, the general direction of the claim—knives exceed rifles—often matches the FBI tables. But the exact “three times” figure is not a constant, and whether it’s accurate depends on which year’s FBI table you are using and on how much “unknown firearm type” is present in that year’s data.
This is where the category definitions matter. A typical pattern in FBI weapon tables is:
– A large number of homicides are attributed to handguns.
– A meaningful number are attributed to firearms where the type is not specified.
– Rifles make up a comparatively small share of firearm homicides.
Against that backdrop, it is not surprising that a broad “knives/cutting instruments” bucket can be several times larger than the narrowly defined “rifle” bucket in a given year.
What the comparison does—and does not—tell you
What it tells you:
– If the FBI weapon table for the year shows more knife homicides than rifle homicides, then the claim “knives kill more people than rifles” is supported for that dataset and year.
– If the ratio is around three-to-one for that year’s counts, then the more specific “three times more” phrasing is also supported.
What it does not tell you:
– It does not mean knives are a larger cause of homicide than firearms. Rifles are only one firearm type; handguns and “unknown firearm type” are substantial categories.
– It does not measure lethality per attack, only counts of homicide incidents by weapon category as recorded.
– It does not settle policy arguments by itself, because it compares a broad cutting-implement bucket to a narrow firearm subset and omits the largest firearm categories.
A more honest way to phrase it
If you want a statement that stays faithful to how the FBI categorizes weapons, it’s better to say something like:
– “In the FBI’s homicide weapon tables, the number of killings attributed to knives or cutting instruments is often higher than the number attributed specifically to rifles.”
And if you want to discuss guns as a whole, the correct comparison is knives versus all firearms, or knives versus handguns, not knives versus rifles.
Bottom line
Using the FBI’s recent homicide weapon breakdowns, the claim can be directionally correct because knife homicides often exceed rifle homicides. But the “three times” ratio is year-dependent, and the comparison is easy to misinterpret because “rifles” excludes handguns and excludes incidents where the firearm type wasn’t recorded. To understand what the FBI data really says, you have to read the weapon categories as the FBI defines them—and resist swapping “rifles” in your head for “guns.”