This study investigates issues of changing recreation use and management in the Greenstone and Caples Valleys. Its underlying research themes are the differing characteristics and activities of four distinct user-groups, how such differences could contribute to conflict in recreation use and management, and the implications of such for management. The conceptual framework used to deal with these issues is the Recreation Opportunity Spectrum (ROS), which is based upon the assumption that quality in recreation experiences and management is best achieved through provision of a range of recreation opportunities. This recognises that objective research input into recreation management cannot substitute for the ultimate subjectivity required in decision-making. Management has two roles here. First, to maintain and/or enhance the range of recreation opportunities availible. Second, to provide users with appropriate signals regarding the availibility of opportunities, and the acceptable norms of use and behaviour. Thus user choice of area and activities undertaken within becomes more a consequence of management action. This differs from most recreation management to date, which has tended to be in response to changing conditions (eg'demand-driven'). The absence of an equitable management approach results in selective reduction of opportunities for certain experiences, as evident from research into effects of conflict/crowding perceptions. In contrast, approaches such as the ROS emphasise management for such opportunities. On this basis and from research results, this study found that maintenance of experiences associated with angling and hunting opportunities in the study area, should be the basis for its management. These opportunities were exploited by relatively more experienced participants. For anglers in particular this was reflected by their characteristics of high activity specialisation. Greater experience and specialisation involved more specific resource requirements and norms of appropriate behaviour. Thus opportunities for hunting and angling in the study area were more susceptible to negative impacts from other uses and users. These impacts would be greatest for angling experiences. This would be more a consequence of perceived inappropriate behaviour by less specialised anglers than a consequence of physical crowding. For hunters these impacts would be primarily a result of concern about the increased presence of others on hunting management, rather than their actual presence. Experiences from the tramping and commercial guided walking opportunities availible are also important, but their participants were relatively less experienced and their recreation opportunities less unique in the region. In the context of an equitable regional ROS approach, there is greater flexibility for their provision elsewhere than there is for experiences from hunting and fishing opportunities. For maintenance of a spectrum of opportunities in the regional ROS, these hunting and angling opportunities should be emphasised in management decision-making. Within the study area itself, management for a regional ROS requires that the Greenstone and Caples Valleys be managed differently. It was clear that the Caples was perceived as providing more 'back-country'-type experience opportunities. In the regional ROS context, management should emphasise maintenance of such opportunities there. Along with this should be noted the greater preference for the CapIes by hunters, and for the Greenstone by anglers.