The objective was to assess effects of experimentally induced undernutrition on responses to an intramammary lipopolysaccharide (LPS) challenge in early-lactation cows. Starting at 24 +/- 3 d in milk, multiparous Holstein cows either received a ration containing 48% straw for 96 h to restrict nutrient intake (REST, n = 8) or were allowed ad libitum intake of a lactation diet (CONT, n = 9). After 72 h on diet or after an equivalent period for CONT, 50 mu g of LPS (Escherichia coli 0111:B4) was injected into one healthy rear mammary quarter to induce an acute inflammation response. Blood samples were collected weekly until 7 wk of lactation, daily during feed restriction (or control), before and at 1, 2, 4, 6, 10, and 24 h relative to LPS injection. Foremilk quarter samples were collected before and at 4, 6, 10, and 24 h after LPS injection. Dry matter intake, milk yield, energy balance, plasma glucose, nonesterifled fatty acids (NEFA), and beta-hydroxybutyrate (BHB) concentrations did not differ between CONT and REST immediately before nutrient restriction in REST (least squares means at d -1 were 21.8, 39.0 kg/d, -2.5 MJ/d, and 3.78, 0.415, 0.66 mM, respectively) but were significantly altered at 72 h of nutrient restriction (9.8, 28.3 kg/d, -81.6 MJ/d, and 2.77, 1.672, and 2.98 mM, respectively), when the LPS challenge was performed. The rectal temperature increment from baseline values in response to LPS did not differ, but cortisol increment was greater and cortisol response area under the curve (AUC) tended to be greater [202 vs. 122 (ng/mL) x 10 h] for REST than CONT. No treatment differences were observed in foremilk IL-8, IL-1 beta, tumor necrosis factor-alpha, and chemokine (C-X-C motif) ligand 3 concentrations in response to LPS injection. Composite milk somatic cell count per milliliter (6.919 x 10(6) vs. 1.956 x 10(6) cells/mL) and total number of somatic cells secreted in milk per day were greater for REST than CONT during the day following LPS. Plasma glucose, urea, and insulin concentrations increased after the LPS challenge, suggesting establishment of insulin resistance and modifications of glucose metabolism to support acute inflammation in both CONT and REST. Nonetheless, nutrient-restricted cows had delayed plasma insulin and glucose responses to LPS, smaller insulin AUC but greater glucose AUC compared with CONT, despite the limited nutrient availability to sustain an inflammation response. Undernutrition altered peripheral metabolic responses to an intramammary LPS challenge but had limited effects on selected indicators of inflammation response in early-lactation cows.