The stimulation of appetite for solid food is of paramount importance in the financial success of an early calf weaning system. In the present study, two production trials suggested that the replacement of the rolled barley (RB) component of a calf starter diet, with molassed sugar beet pulp (MSBP) or lucerne meal (LM), stimulated appetite, particularly before weaning. Despite continued improvements in voluntary food intake (VFI) after weaning, dietary differences were reduced, with no significant differences observed. The optimum level of MSBP inclusion in a barley based diet appeared to be in the range of 0.15-0.25 of diet DM. All diets bar one containing 0.55 MSBP, produced live weight gains (LWG) in excess of 0.7kg/d. However, the relationship between LWG and VFI was considerably stronger for barley based diets than those based on MSBP or LM. Metabolic studies suggested two possible explanations for the reduced LWG response to VFI as MSBP or LM replaced RB: a) inefficiency in ruminal fermentation and/or b) reduced nitrogen (N) digestibility. The reduction in N digestibility was correlated with an increased proportion of N bound to the ADF matrix. Further studies suggested that a microbial population capable of complete cell wall degradation was absent in the rumens of young calves up to 3 months of age. The results suggested that the availability and quality of the N leaving the rumen was reduced as MSBP or LM replaced RB. It was unclear from the present studies whether increased diet buffering capacity or a preference for feedstuffs other than RB could explain the potential improvements in VFI. A method was derivised to measure the buffering capacity in the rumen fluid of young calves which showed that the total VFA are the principal chemical component of the buffering system and the only rumen parameter found to make significant contributions to mathematical models accounting for variation in VFI. Rumen pH, although low, was not a regulator of VFI and accounted for only 0.03 of VFI variation.