Humans have used wild furbearers for various purposes for thousands of years. Today, furbearers are sustainably used by the public for their pelts, leather, bones, glands, meat, or other purposes. In North America, contemporary harvest of furbearers has evolved along with trap technologies and societal concerns, and is now highly regulated and more closely coupled with harvest analysis and population monitoring. Traps and regulated trapping programs provide personal or cultural rewards that can also support conservation, and can assist with advancing ecological knowledge through research, protecting endangered species, restoring populations or habitats, protecting personal property, and enhancing public health and safety. However, animal welfare and trap selectivity remain important topics for furbearer management in North America, as they have for more than a century. A related international challenge to modern furbearer management came with the Wild Fur Regulation by the European Union, which passed in 1991. This regulation prohibited use of foothold traps in many European countries and the importation of furs and manufactured fur products to Europe from countries that allowed use of foothold traps or trapping methods that did not meet internationally agreed-upon humane trapping standards. To address existing national concerns and requirements of the Wild Fur Regulation, the United States and European Union signed a non-binding bilateral understanding that included a commitment by the United States to evaluate trap performance and advance the use of improved traps through development of best management practices (BMPs) for trapping. Our testing followed internationally accepted restraining-trap standards for quantifying injuries and capture efficiency, and we established BMP pass-fail thresholds for these metrics. We also quantified furbearer selectivity, and qualitatively assessed practicality and user safety for each trap, yielding overall species-specific performance profiles for individual trap models. We present performance data for 84 models of restraining traps (6 cage traps, 68 foothold traps, 9 foot-encapsulating traps, and 1 power-activated footsnare) on 19 furbearing species, or 231 trap-species combinations. We conducted post-mortem examinations on 8,566 furbearers captured by trappers. Of the 231 trap model-species combinations tested, we had sufficient data to evaluate 173 combinations, of which about 59% met all BMP criteria. Pooling species, cage traps produced the lowest average injury score (common injuries included tooth breakage), with minimal differences across other trap types; species-specific patterns