Land of the semi-arid zone of Australia is generally managed to produce wool or beef. Past management has caused many changes in the land. These changes may be difficult to detect and assess. Much of the available information is at too coarse a scale to be really useful in assessing change. Graziers' perceptions of change are unknown but survey results from the agricultural zone suggest that their perceptions are probably incorrect. Apportioning the causes of change is very difficult as the main agents (climate, stocking rates, bushfires, legislation and economics) are not independent. Three different approaches to separating cause are described: use of historical information, integrating all information and using unpalatable plants as proxies for key economic species. Some difficulties with the historical approach are outlined. The major issues in semi-arid land management are social rather than technical. However, key aspects such as perceptions, motivation, and sources of information used by graziers are neglected research subjects. Recent research into an objective basis for assessing stocking rates from forage biomass production will replace traditional estimates based on extrapolating from similar country. This will significantly assist graziers in determining appropriate stocking rates to maximise their incomes. Other research by graziers has demonstrated the benefits of low stocking rates leading to increased incomes on both an animal and area basis. Such advances by graziers provide keys for future extension programs to achieve the desired goals of a stable grazing industry with good financial rewards, and improved land management.