Maximisation of aggregate recreational hunter benefits involves managing both the prey and the hunter. The biology of game animals, and hence the supply side of the management situation, is reasonably well understood, but there is relatively little information on the demand side. On public lands, where there is no market to signal the quality of the hunting experience the game manager has little guidance on how to allocate the resource amongst individual hunters. In New Zealand, there is no attempt to do so. While seeing and killing game are known to enhance individual hunters’ benefits, the allocation of the resource across hunters raises the prospect of limiting individual hunter harvests, normally enacted through a bag limit. The benefits of doing so are dependent upon the marginal benefits of harvest for different hunters. The relationship between hunter satisfaction and the number of animals killed is explored using data from a longitudinal study of a large group of deer hunters. Latent class models of satisfaction outperform random parameters models and identify heterogenous groups of hunters whose satisfaction is differentially dependent on game sightings and harvest. Personal attributes and hunter motivations help explain some of these differences. Heterogenous and rapidly diminishing marginal satisfaction present a strong case for management of at least part of the open-access New Zealand red deer herd to enhance social welfare by increasing the number of hunters harvesting a deer rather than going home empty-handed.