Abstract: A ground cover index (GCI) derived from Landsat TM data is available for Queensland since 1986. It provides a potentially robust means for reporting change in landscape function to the Australian Collaborative Rangelands Information System (ACRIS). We report early progress in developing such an index based on the temporal behaviour of pixel GCI values. Pixels within paddocks in the Wambiana grazing trial were classified as notionally ‘functional’ or ‘dysfunctional’ based on a specified deviation in ground cover from their longer-term temporal mean. This produced distinct spatial patterning of ‘dysfunctional’ areas for some grazing treatments in drier years. The proportional area of this ‘dysfunctional’ class within paddocks and vegetation types was significantly correlated with estimated ground cover. Further methodological development will relate the currently developed temporal deviations in pixel GCI to that occurring in the surrounding neighbourhood. When a suitably robust method for indicating landscape function from GCI is available, we will increase the scale of analysis to sub-bioregions so as to report change in landscape function for Queensland’s rangelands to ACRIS.