Two child restraint models were subjected to a series of 30-mph sled tests, using the FMVSS 213 standard procedures and the 3-year-old child dummy, except that the restraints were mounted on production vehicle seats. Lap belts were anchored at various points on a 5-in radius, 105° arc. All but one anchorage location was within the range allowed by FMVSS 210. It was concluded that child restraint structural integrity and forward excursions are degraded as belt angles become steeper and the anchorage location approaches and goes forward of the vehicle seat bight line. Recommendations are made for (a) a longitudinal limit on belt anchorage location relative to a measuring fixture developed by the SAE Children's Restraint Systems Task Force and (b) consideration by child restraint manufacturers of the effects of different belt angles on child restraint performance.