Although they are used frequently in veterinary dermatology, the reliability of canine pruritus severity scales has not been reported. The objective of this study was to evaluate the reliability of pruritus severity numerical rating scales (NRS) and pruritus severity visual analogue scales (VAS). Videos of 16 dogs were evaluated for pruritus severity by 24 observers utilizing three NRS and three VAS. Intraobserver repeatability and interobserver reproducibility were evaluated with Cohen's kappa and Kendall's rank correlation statistics, respectively. The repeatability of pruritus severity NRS was fair, with mean Cohen's weighted kappa (κw) values ranging from 0.49 to 0.60. The mean Kendall's rank correlation coefficient (t) for the three VAS ranged from 0.62 to 0.73. The reproducibility of mid-range pruritus severity ranks was poor with both scale types. Scales describing overall pruritus severity were found to be reliable most consistently. Neither NRS nor VAS displayed the degree of reliability desired in a health measurement scale. Interpretation of research results evaluating canine pruritus severity with NRS and VAS scales should account for suboptimal reliability.