Fractal geometry is a useful tool for the analysis of landscape data. In this study fractal scaling was applied to high-resolution landscape data collected with a profiling laser altimeter. The objective of this work was to assess the persistence of scaling differences over time. Data were collected at the United States Department of Agriculture (USDA), Agriculture Research Service (ARS) Jornada Experimental Range in New Mexico, USA in May and September 1995 and February 1996 over a grass-dominated site, a shrub-dominated site, and a transitional area between shrub- and grass-dominated sites along four transects at each site for each date. Root-mean-square (RMS) roughness was scaledependent and had more than one range of self-affine scaling. Different numbers of self-affine scaling intervals, boundaries of intervals, and fractal dimensions over these intervals were associated with different land covers. A linearity measure was applied to find intervals of fractal scaling. The number and boundaries of fractal scaling intervals appeared to be persistent over the year. Grass and shrub sites had two and four linearity intervals respectively. The transitional site had a pattern of scaling that was intermediate between grass and shrub sites. The lowest fractal dimensions at small scales of 6-30 m corresponded to the maximum vegetation in September.