Two manufacturers, Denton ATD and FTSS, currently produce the Hybrid III 5th percentile female dummy. In response to concerns raised by industry that differences in the anthropometry of the molded breasts between the two manufacturers may influence chest responses, Transport Canada conducted a comparative testing program. Thorax biofidelity tests were conducted to compare force-deflection characteristics; full-frontal, rigid-barrier tests were conducted at 40, 48 and 56 km/h to compare chest responses, and out-of-position chest on module static airbag deployment tests were conducted to compare peak chest deflections of the Denton and FTSS dummy jackets and of a prototype jacket without breasts.Differences in force-deflection characteristics were observed during biofidelity pendulum impacts of the two dummies, with much of the differences attributed to the different chest jackets. Differences of up to 11 mm in the peak sternum deflection and of the order of 15 g for the 3-ms chest acceleration clips were noted in rigid-barrier vehicle tests. In the out-of-position airbag deployment tests, differences in the magnitude of peak chest deflections were observed. The prototype chest jacket without breasts was found to improve repeatability in the belted crash tests and in out-of-position airbag testing.Though neither the Denton nor the FTSS chest jackets fully meet the original design intent of the Hybrid III 5th percentile dummy, the Denton dummy more closely met the drawing specifications and had less manufacturing variability. The results demonstrate the importance of detailed chest flesh assembly specifications, provide evidence that a fully molded jacket design would eliminate manufacturing variability and suggest that removal of the breasts may further improve test repeatability.