AbstractLead bullet fragments pose a health risk to scavengers and hunters consuming game meat, but lead or lead-core bullets are still commonly employed for big and small game hunting. Bullet fragmentation has been assessed for modern, high-velocity rifles, but has not been well documented for black-powder cartridge rifles or muzzleloading firearms. We used two established methods to estimate bullet fragmentation. We evaluated a traditional .54 round ball and a modern-designed .54 conical bullet for muzzleloaders, two types of .45-70 black powder rifle cartridges, and a modern lead-core high-velocity bullet (.30-06) as our comparison control. We tested penetration and fragmentation in water (n = 12) and ballistics gel (n = 2) for each bullet type. We measured lead mass lost to fragmentation and x-rayed ballistic gels to visualize fragmentation patterns. The modern .30-06 bullets we tested (Remington Core-Lokt) retained a mean of only 57.5% of original mass, whereas mean retention by muzzleloader and black powder cartridge bullets ranged 87.8-99.7%. Round balls and .45-70 bullets shed less lead (i.e., 0.04g and 0.19g on average respectively) than the modern conical .54 muzzleloading bullets (3.08g) or the .30-06 control (4.14g). Fragments from round balls and black powder cartridge bullets showed far less lateral spread compared to the high-velocity modern bullet. Our findings suggest that round balls for muzzleloaders and black powder cartridge bullets may leave far fewer lead fragments in game than the conical muzzleloader bullet or modern high-velocity rifle bullet we tested, and thus could pose a lower risk of secondary lead poisoning for humans and wildlife. Artificial tests cannot replicate conditions encountered in the field, but the striking differences we observed in bullet fragmentation even under severe testing conditions suggests that follow-up tests on game animals may be warranted.