Summary Reasons for performing study The mechanism of hyperthermia, a potentially fatal adverse effect of erythromycin treatment of foals, is unknown. Objectives To determine the cause of erythromycin‐associated hyperthermia. It was hypothesised that the normal sweat response of foals is impaired by treatment with erythromycin. Study design Blinded, crossover study in 10 healthy pony foals. Methods Foals kept in stalls were given either erythromycin (25 mg/kg bwt orally, 3 times daily) or control for 10 days then turned out for a further 10 days. Quantitative intradermal terbutaline sweat tests were performed on Days 1 (baseline), 3, 10 and 20. The effects on terbutaline‐induced sweating of erythromycin, terbutaline concentration and treatment day were analysed by repeated‐measures ANOVA with Bonferroni‐corrected pairwise post hoc comparisons. Peak temperatures were compared by Wilcoxon's signed rank test and proportions by McNemar's related samples test. Significance was set at P<0.05. Results There were significant 2‐factor interactions for treatment × terbutaline after baseline, treatment × day at every terbutaline concentration, and day × terbutaline for erythromycin (P<0.001) but not control (P = 0.9) treatment. Sweating was significantly reduced from baseline in erythromycin‐treated foals at all subsequent days. Erythromycin‐treated foals produced less sweat at all time‐points than did control‐treated foals (P<0.05). Peak rectal temperatures of erythromycin‐treated foals were significantly higher (P = 0.02) than those of controls. During the first 3 days outside more erythromycin‐treated than control‐treated foals required treatment for hyperthermia (6 vs. 0; P = 0.03). Conclusions We believe drug‐induced anhidrosis is the likely cause of hyperthermia in some foals treated with erythromycin.