A series of sled tests was performed to analyze the responses of an anthropomorphic test device (ATD), particularly neck forces, when rear-facing child restraint systems (CRS) are tethered. Nominally identical rear-facing CRS were tested in four tether conditions: untethered, tethered down to the floor, tethered down to the bottom of the vehicle seat, and tethered rearward to a point above the back of the vehicle seat. The CRABI 12MO ATD with head, upper neck, and chest instrumentation was used in all tests. The tests were conducted using the ECE R44.02 test bench. Both frontal and rear impacts were performed and each condition was repeated for a total of 16 sled tests. Motions of the CRS and ATD were recorded using high-speed digital video (1000 fps). The highest ATD accelerations, forces, and moments were observed during the primary impact of a frontal test, rather than on rebound. The loads observed during rebound from frontal impact were similar in magnitude to the peak loads collected during rear impact. The four tethering geometries produced distinct loading patterns. The lowest HIC, neck forces, and chest accelerations in both impact directions were observed with the rearward tether. The upper neck moment data did not show a clear trend relative to tethering geometry. ATD and CRS motions were best controlled in frontal impact by the rearward tethering geometry while the motions in rear impact were best controlled by tethering to the floor. The data show a potential benefit in both frontal and rear impacts of tethering rear-facing CRS to a point above the vehicle seatback.