Accident data show that the injury risks to children seated in Child Restraint Systems (CRSs) are higher in side collisions than any other type. To investigate this, it is necessary to understand the occupant responses in various CRSs. While a sled test procedure has been discussed by the International Organisation for Standardisation for evaluating CRS performance, for which full car side impact tests were conducted with a Q3s child dummy seated in a Forward-Facing (FF) CRS, and a CRABI Six-Month-Old (6MO) infant dummy seated in a Rear-Facing (RF) CRS and also placed in car-bed restraint. In FF and RF CRS tests, the chest was loaded by an intruding door and the resulting chest deflection of the Q3s child dummy in the FF CRS was comparable with the Injury Assessment Reference Value (IARV). In tests with the FF and RF CRSs, the dummy heads were contained within the CRS shell during the entire impact event; and the HICs were small. Although all injury criteria of the CRABI 6MO infant dummy in the car-bed restraint were smaller than the IARVs, a calculated crotch harness force was large.