This study was aimed at improved understanding of the mechanisms of previously reported protective effects of a pneumatic boxing glove. A Motion Capture System was used to obtain velocity data from four different boxing gloves dropped on to a force plate from nine heights ranging from 1 to 5 metres. Two gloves were of the conventional type but differed in mass. The other two were prototype pneumatic gloves. One of these (SBLI) had a sealed bladder while the other (ARLI) incorporated a port allowing air exchange with the external environment. The pneumatic gloves decelerated more slowly than the conventional gloves following impact and compressed through a greater absolute distance. Consequently, they took longer to reach zero velocity. As drop height increased, these trends became more pronounced for the ARLI glove than the SBLI glove. Increase in velocity during rebound was also slower for the pneumatic gloves. The ARLI glove had a lower coefficient of restitution than any of the other gloves at low to moderate drop heights but not at high drop heights. The SBLI glove had a higher coefficient of restitution than the other gloves at all drop heights from 2 metres upwards. This indicated that, overall, the ARLI glove was the most effective, and the SBLI glove the least effective, in dissipating the kinetic energy of impact through conversion to other energy forms. For all gloves at all drop heights, peak positive acceleration at the beginning of rebound was of lower absolute magnitude than peak negative acceleration at the end of compression. The influence of drop height on an index characterising this relationship differed between the conventional and pneumatic gloves, possibly reflecting structural changes to gloves as impact energy increased. The conventional and pneumatic gloves differed regarding temporal alignment between key kinematic and kinetic events, and there were some differences between the two pneumatic gloves in this respect. Nevertheless, peak glove deceleration correlated highly with peak impact force, not only for each glove individually but also when data for all gloves were combined. The findings confirmed the potential practical utility of the ARLI glove and identified air cushion thickness, glove compressibility and capacity for air release and subsequent reuptake as critical aspects of its design.