Measurements of the transmission of vibration from the floor beneath a seat to interfaces between the seat and the human body are used to characterise the dynamic performance of the seat. Most studies of seat transmissibility concern the transmission of vertical vibration through seat pan cushions, but fore-and-aft vibration at a seat pan and backrest can also cause discomfort. This study investigated the transmission of fore-and-aft vibration through a car seat to surfaces on the seat pan cushion, the backrest cushion and the headrest, and also to the seat frame at these three locations. The study sought to understand the influences of the seat frame and the cushion foam and to provide data for dynamic modelling of the seat. The fore-and-aft transmissibility of the seat was measured with 12 human subjects and an SAE J826 manikin, using 120-s periods of fore-and-aft random vibration (0.25-40 Hz) at three magnitudes (0.4 m/s 2 r.m.s., 0.8 m/s 2 r.m.s. and 1.2 m/s 2 r.m.s.). With the manikin, there were three resonances (at 3.5 Hz, 12 Hz and 20 Hz) in the transmissibilities from the seat base to the seat pan frame and the seat pan cushion, and to the backrest frame and the backrest cushion. With the human subjects, the transmissibilities from the seat base to all six locations on the seat showed a principal resonance around 4 Hz. With increasing magnitude of the vibration, the principal resonance frequency in all the seat transmissibilities decreased. It is concluded that fore-and-aft resonances in the seat transmissibilities, especially to the backrest, are likely to affect vibration discomfort. There are large differences between the foreand-aft transmissibility of a seat with a manikin and those with human subjects, consistent with the human body having dynamics very different from those of a rigid mass. Non-linearities in the seat transmissibilities with the manikin and with the human subjects may be explained by non-linearity in the biodynamics of the body and the responses of the seat and the manikin.