Cow–calf operations constitute the main land-use system in the 18 million ha of well drained tropical savannas of Colombia located in the Orinoco basin. There, numerous sex and age animal categories are present, ranging from suckling calves to old cull cows and steers, which in variable proportions are stocked at one animal unit (AU)/5–10 ha, depending on the ranch and distance from markets. In parallel, early-on farm observations showed that when sown pastures are introduced, graziers recur to opportunistic partial fattening of cull animals for a rapid economic return. The paper reports animal weight gains of cull cows, old steers, and young heifers and yearlings over seven consecutive years grazing on a low fertiliser-input well managed Andropogon gayanus-based pasture and four contemporary years on Brachiaria humidicola. The first 2 years involved three stocking rates (SR; 1.38, 1.85 and 2.32 animal units/ha), whereas a single SR of 1.33 animal units/ha was used for the remainder. A large database that included chemical, and in vitro and in vivo digestibility, values for all forages involved, was used as an input to a mathematical model to estimate methane (CH4) emissions. Estimations were compared with weight-based predictions derived from chamber measurements recorded recently in tropical Australia. Faecal mineral concentration varied little (not significantly) among SR and animal categories. Weight gains of old steers were 4–24% larger than those of cull cows, which, in turn, were larger than those of young steers and heifers (P < 0.01). Excepting one year, the performance of cull cows and old steers did not differ between pastures, confirming earlier on-ranch observations of reasonable weight gains of adult cattle on both pastures. The predicted carcass composition of cull cows and old steers did not differ between pastures either. Over the length of the experiment, cull cows emitted significantly less CH4 than did old steers (129 vs 141 g/day) on A. gayanus, while emissions on B. humidicola amounted to 128 and 137 g/day respectively. Despite between-year differences in animal performance, they are indicative of well managed pasture sustainability. They also show some of the trade-offs involved in the use of contrasting animal categories and pastures in terms of weight performance and predicted carcass composition and CH4 production. The authors expect that the present results will contribute to a rational, science-based discussion of the role of beef systems vis-à-vis environment in a region with limited production alternatives.