The hypotheses that endophyte (Neotyphodium coenophialum)-infected tall fescue (TF) seed causes vasoconstriction in horses in vivo and that ground seed would cause more pronounced vasoconstriction than whole seed were tested. Ten horses each received 1 of 3 treatments: endophyte-free ground (E-G; n = 4 horses) seed, endophyte-positive whole (E+W; n = 3) seed, or endophyte-positive ground (E+G; n = 3) seed. There were two 14-d periods, P1 and P2. During P1, animals were adapted to a concentrate (0.2% BW, as fed, twice daily) and alfalfa cubes. During P2, the seed was mixed into the concentrate portion of the diet and alfalfa cubes were offered ad libitum. Fescue seed was fed in increasing amounts ranging from 0.02% BW on d 1 (averaging 76 ug/kg ergovaline + ergovalinine) to 0.22% BW on d 11 to 14 (averaging 713 ug/kg ergovaline + ergovalinine). The distal palmar artery of the left foreleg of each horse was scanned via Doppler ultrasonography for 4 d during each period, with 5 replicate scans performed on each scanning day. The measurements taken at each scan included artery luminal diameter, area, and circumference, peak systolic velocity, end diastolic velocity and blood flow variables. Animal temperature, heart rate, and respiration rate and ambient temperature and humidity were also recorded. Blood samples were taken on each scanning day to measure inflammatory cytokine mRNA abundances, and blood samples were collected on d 0, 4, 8, and 14 of P2 to measure prolactin concentrations. Consumption of E+G TF seed caused decreased artery lumen diameter (P = 0.0033), area (P = 0.0406), and circumference (P = 0.0480) compared with E-G seed, and E+W seed produced an intermediate response. Blood flow volume was reduced (P < 0.05) during P2 in horses receiving E+G seed compared with horses receiving E-G seed. Other ultrasound variables were not different (P > 0.05) among treatment groups, and neither were cytokine mRNA or prolactin concentrations. Treatment did not alter (P > 0.05) animal temperature, heart rate, or respiration rate, and neither ambient temperature nor relative humidity was consistently correlated with any response variable measured. Taken together, these data confirm that consumption of E+G fescue seed caused vasoconstriction in horses, which could be readily measured by Doppler ultrasonography. Use of Doppler ultrasound to monitor the diameter of the palmar artery of horses grazing endophyte-infected (E+) fescue pastures may provide a convenient and satisfactory biomarker to determine premonitory signs of fescue toxicosis.