Despite all the biotic and abiotic factors affecting foraging by ruminants, there is a common and fundamental process, which is bite gathering. We hypothesised that because the mechanics of bite formation dominate the foraging process, changes in short-term bite mass are reflected in longer-term animal performance across a wide range of sward conditions. We focus at the meal level of foraging, using experiments in which the effect of abiotic factors and digestive constrains are minimised, making intake rate the main currency. We estimated bite mass across a wide range of structural challenges to large-herbivore foraging in a long-term experiment with heterogeneous native grasslands. A conceptual model was developed for average daily gain, where energy gain and energy costs were proximate causal variables. Energy gain was a function of diet quality and components of daily intake rate, where bite mass was the main component estimated. In turn, components of intake rate were determined by sward structure and bodyweight. Energy costs were a function of bodyweight and abiotic conditions. Finally, sward structure, bodyweight and abiotic conditions were determined by experimental treatments, seasons and years. Then, the conceptual model was translated into statistical models that included variables measured or estimated, and coefficients representing all links in the conceptual model. Weight gain was a function of bite mass, forage characteristics, and animal and abiotic conditions. Models were set up to test whether forage and stocking conditions affected monthly gain beyond the effects through bite mass, after correcting for abiotic factors. Forage mass, height and disappearance did help predict monthly gain after bite mass was included in the model, which supported our hypothesis. However, stocking treatments and season had significant effects not incorporated in bite mass. Although the model explained 77.9% of liveweight gain variation, only 35.2% was due to fixed effects, with 10.8% accounted by bite mass and its interactions. Concomitant experiments showed that sward structure (first with sward height and the second with tussock cover) does determine bite mass and short-term intake rate in the complex native grasslands we studied. Yet, other temporal varying components of monthly gain not correlated with bite mass, temperature or wind, added most of the observed variation in monthly animal performance. Part of the model failure to account for variation in performance may be related to a significant and temporally variable grazing of tussocks. We used a bite mass model that assumed no tussock grazing. In light of these results and a parallel experiment, we conclude that tussock grazing must be incorporated in future versions of the model.