A comprehensive review of casualty-reducing effectiveness estimates of child safety seats in actual use, obtained by statistical analyses of highway accident data. Recent analyses of large samples of New York and Maryland accidents show statistically significant injury reductions for child safety seats; so does a new analysis of the National Highway Traffic Safety Administration's accident files. Results from Washington State, Tennessee, New Jersey, and Idaho are also reviewed, as are Nationwide restraint usage and fatality trends. The findings are critically examined for possible data biases. It is concluded that child safety seats definitely reduce deaths and injuries in highway crashes, but that their effectiveness cannot be accurately estimated at this time because of inconsistencies and possible biases in the various studies.