Objective - To investigate the effects of dexamethasone or levothyroxine sodium on endotoxin-induced alterations in glucose and insulin dynamics. Animals - 24 horses. Procedures - Horses were randomly allocated to 3 treatment groups and received 48 mg of levothyroxine mixed with 200 g of oats, 20 mg of dexamethasone plus oats, or oats alone (control) for 15 days, followed by IV infusion of lipopolysaccharide (20 ng/kg) while individually housed in stalls. Frequently sampled IV glucose tolerance tests were performed prior to pretreatment, after pretreatment, and 20 hours after lipopolysaccharide administration. Area under the curve for plasma glucose and serum insulin concentrations was calculated, and minimal model analyses were performed. Results - Significant treatment-by-time effects were detected for insulin sensitivity (SI) and area under the curve for glucose and insulin in the 15-day pretreatment period. Insulin sensitivity significantly decreased over time in all treatment groups, with the largest decrease detected in the dexamethasone group. Administration of lipopolysaccharide further decreased mean SI by 71% and 63% in the dexamethasone and control groups, respectively, but did not affect horses in the levothyroxine group. Mean SI was the lowest in the dexamethasone group, but percentage reduction was the same for dexamethasone and control groups. Conclusions and Clinical Relevance - Insulin sensitivity decreased during the pretreatment period in all 3 groups, indicating that hospitalization affected glucose and insulin dynamics. Dexamethasone significantly lowered SI, and endotoxemia further exacerbated insulin resistance. In contrast, there was no additional effect of endotoxemia on SI in horses pretreated with levothyroxine, suggesting that this treatment prevented endotoxemia-induced insulin resistance.