Physical rehabilitation therapy is rapidly becoming an integral component in the treatment of a variety of neurologic and orthopedic conditions in veterinary patients. However, to date there are no perceived exertion scales (PES) available for monitoring dogs or for developing exercise protocols. The purpose of this study was to develop a perceived exertion scale for dogs walking/running on a land treadmill using the following physiologic parameters: heart rate, respiratory rate, rectal temperature, blood lactate and glucose concentrations, and regional tissue oximetry. We hypothesized that the measured physiologic parameters would correlate with the perceived exertion of the dogs exercising on a land treadmill and that there would be no difference between intra-and inter-observer perceived exertion of dogs while walking/running on a land treadmill. An additional aim was to compare heart rate obtained from a Holter monitor to those obtained manually and using a mobile ECG monitor. Fifteen healthy client-owned dogs were enrolled in this prospective clinical trial. Each dog participated in two 55-minute exercise trials separated by a one-week recovery period. An initial complete blood count, biochemical panel, urinalysis, total T4, and venous blood gas (including lactate) were performed prior to beginning each trial. Subsequent venous blood gases were obtained after 20 minutes of exercise, immediately after completion of exercise, and after a 20-minute recovery period. Additional physiologic parameters recorded included heart rate, rectal temperature, respiratory rate, and regional tissue oximetry. The perceived exertion was recorded during each trial at 2 minute intervals by a Certified Veterinary Technician, who is also a Certified Canine Rehabilitation Practitioner. Additional perceived exertion scores were recorded via video recording by two veterinarians, a rehabilitation therapy volunteer, and a iii veterinary student to determine inter-observer variation. The results of the first trial were also compared to the second trial to evaluate for repeatability. Thirteen of 15 dogs completed both trials. Spearman correlations for both trials were positive between Holter heart rate and KardiaMobile heart rate (r = 0.7, p <0.0001), manual heart rate (r = 0.64, p <0.0001),), respiratory rate (r = 0.27, p = 0.019), and temperature (r = 0.29, p = 0.014). Holter heart rate was positively correlated with PES scores from all observers for both trials (p <0.001 for all). The mean PES score recorded by each observer was 2 out of 4 for both trials. To test the assumption that physiological measurements changed with PES score, each dog's time 0 value was subtracted from subsequent values for that dog and least trimmed squares regression was then used to remove outliers. Outlier-free linear regression estimates were then obtained as follows. For Trial 1 glucose (g/dL) =-2.581-2.112 PES (R 2 = 0.179, F(1,45) = 9.787, P=0.003). That is, glucose decreased 2.112 g/dL for each incremental increase on the PES. For Trial 2 cutaneous ox...