Long-term information on the effects of managed grazing versus excluded grazing effects on vegetation composition of desert rangelands is limited. Our study objectives were to evaluate changes in frequency of vegetation components and ecological condition scores under managed livestock grazing and excluded livestock grazing over a 38-yr period at various locations in the Chihuahuan Desert of southwestern New Mexico. Sampling occurred in 1962, 1981, 1992, 1998, 1999, and 2000. Range sites of loamy (1), gravelly (2), sandy (2), and shallow sandy (2) soils were used as replications. Black grama (Bouteloua eriopoda Torr.) was the primary vegetation component at the seven locations. Dyksterhuis quantitative climax procedures were used to determine trends in plant frequency based on a 1.91-cm loop and rangeland ecological condition scores. Frequency measures of total perennial grass, black grama, tobosa (Hilaria mutica Buckley), total shrubs, honey mesquite (Prosopis glandulosa Torr.), and other vegetation components were similar on both grazed and ungrazed treatments (P > 0.1) at the beginning and end of the study. The amount of change in rangeland ecological condition scores was the same positive increase (14%) for both grazed and ungrazed treatments. Major changes (P < 0.1) occurred within this 38-yr study period in ecological condition scores and frequency of total perennial grasses and black grama in response to annual fluctuations in precipitation. Based on this research, managed livestock grazing and excluded livestock grazing had the same long-term effects on change in plant frequency and rangeland ecological condition; thus, it appears that managed livestock grazing is sustainable on Chihuahuan desert rangelands receiving over 25 cm annual precipitation.