Cow–calf productivity on 2 lightly (25%–30% use) and 2 conservatively grazed pastures (35%–40% use) were evaluated over a 5-year-period (1997 to 2001) in the Chihuahuan Desert of south-central New Mexico. Spring calving Brangus cows were randomly assigned to study pastures in January of each year. Experimental pastures were similar in area (1 098 ± 69 ha, mean ± SE) with similar terrain and distance to water. Use of primary forage species averaged 28.8% ± 4.3% in lightly stocked pastures and 41.8% ± 4.4% on conservatively grazed pastures. Perennial grass standing crop (168.8 ± 86 vs. 173.6 ± 58.3 kg·ha⁻¹) and adjusted 205-day calf weaning weights (279.1 ± 7.5 vs. 270.7 ± 7.8 kg) did not differ among lightly and conservatively grazed pastures. Cow body condition scores in autumn, winter, and spring were similar among grazing levels as were autumn and winter body weights. However, cow body weights tended to be heavier (P < 0.10) in lightly grazed pastures relative to conservatively grazed pastures (524 vs. 502 ± 9.7 kg) in spring. Lightly grazed pastures yielded greater (P < 0.05) kg of calf weaned·ha⁻¹ and calf crop percent than conservatively grazed pastures in 1998 due to destocking of conservatively grazed pastures during that year's drought. Conversely, pregnancy percent tended to be greater (P < 0.1) in conservatively relative to lightly grazed pastures (92.6% vs. 87.7%); however, this advantage is explained by herd management as cows in the conservatively grazed pastures were removed during drought of 1998, avoiding exposure to the drought stress experienced by cows in the lightly grazed pastures. Nonetheless, pregnancy percents from both grazing treatments would be acceptable for most range beef production systems. Results suggest that consistently applying light grazing use of forage is a practical approach for Chihuahuan Desert cow–calf operations to avoid herd liquidation during short term drought.