The study objective was to evaluate the effects of reducing dietary starch content in fresh cow diets while maintaining NDF levels by substituting barley grain (BG), corn silage (CS), or both with beet pulp (BP) on DMI, lactation performance, serum mineral and metabolites concentrations, liver enzymes and liver functionality index (LFI), serum insulin and revised quantitative insulin sensitivity check index (RQUICKI). Thirty-six multiparous cows were randomly assigned to 1 of 4 experimental diets from calving to 21 days in lactation. Experimental diets were a high-starch diet with ground BG (CO; 24.9% starch; 0% BP) and 3 low-starch diets where BP substituted for either BG (BB; 19.6% starch; 7% BP), CS (BC; 20.6% starch; 12% BP) or CS and BG (BCB; 20.3% starch; 12% BP). Relative to CO cows (16.50 kg/d), DMI was greater for BC (17.70 kg/d) and BCB (17.50 kg/d) cows, but it was lesser in BB (15.60 kg/d) cows. Similar to DMI results, milk yields tended to be greater for BC (37.89 kg/d) and BCB cows (37.81 kg/d) compared to CO cows (35.41 kg/d), but BB cows (33.05 kg/d) tended to produce less milk than CO cows. Relative to CO, cows fed BB had lower serum glucose concentrations, whereas cows fed BC and BCB had higher serum glucose. Serum insulin concentrations were lower for BB cows than for other exprimental groups. Relative to CO, serum NEFA and BHB concentrations were lower for BC and BCB, but was similar between BB and CO. The RQUICKI was lower for CO, BC cows than BB cows, and cows in BCB tended to have less RQUICKI compared to BB during postpartum. The concentrations of gamma glutamyl transferase were lower in BC and BCB cows relative to CO and BB cows. The cows fed CO and BB had higher serum bilirubin relative to cows fed BC and BCB diets. Although, LFI for CO cows was similar to cows on BB, BC and BCB, BC cows had higher LFI than BB cows and tended to have greater LFI than CO cows. Overall, reducing dietary starch by replacing CS (BC) or a mix of CS and BG (BCB) with BP positively affected DMI and milk yield and indicated improved energy metabolism and liver function during the first 21 d of lactation compared to when BP was fed instead of BG to reduce starch (BB).