Pressures to reduce stocking densities (extensify) are arising for a number of reasons. There are two main options once the decision to reduce stocking occurs: (1) to maintain management and physical inputs pro rata to remaining livestock which may or may not have full access to the same land area; (2) to reduce management and/or physical inputs to remaining stock because of real or perceived increases in availability of grazed resources to remaining livestock. These two options have widely different implications for the health, welfare and performance of the extensified systems. A third option open to producers is merely to reduce inputs, de-intensify, without reducing stock numbers. All the options raise major questions in relation to their benefit or otherwise to ecological or landscape values which tend to be underpinning these initiatives. However, significant uncertainty arises as to how producers will respond. Economic factors will play a part in justifying changes and levels of compensation deemed adequate. Levels of animal performance are crucial in terms of economic viability and as an indicator of animal welfare. Three programmes of research work are in hand in the UK which address these concerns and the associated impact on the environment. At ADAS's Redesdale and Pwllpeiran Farms the experimental protocol has been to reduce stocking rate with no change in inputs per ewe. At SAC's Kirkton and Auchtertyre unit a different farm scale strategy has been adopted. Here stock numbers have been reduced, but alongside these physical (food) and management inputs (labour) have also been reduced. Preliminary data on animal performance from all experiments suggests no improvement in animal performance as a result of reduced stocking density. In the latter study, where inputs per ewe have also been reduced, reduced output per ewe has been found, partly as a result of increased lamb mortality.