Crippling loss (waterfowl downed, but not retrieved) is a recognized component of waterfowl harvest, but there are few contemporary estimates. We used a zero-inflated negative binomial model to account for false reporting of zeros and estimated waterfowl crippling losses from harvest data reported in a 2018-2019 survey of waterfowl hunters in Illinois, USA. We found that crippling rates declined rapidly with increasing harvest and moderately with increasing days afield. For the most active hunters (days afield), the probability of falsely reporting zero crippling loss increased with the number of ducks they reported harvesting, whereas for the least active hunters we found the reverse. We suggest that modest improvements in hunter skill could produce significant reductions in crippling loss.