BackgroundTo optimise the interleukin (IL)‐31‐blocking therapy in atopic dermatitis (AD), an understanding of the chronology in the expression of IL‐31 and its receptor (IL‐31RA) is needed.Hypothesis/Objectives(i) To assess the chronological expression of IL‐31 in canine AD skin lesions, (ii) to compare it with serum IL‐31 levels and macroscopic skin lesion scores, and (iii) to determine the identity of IL‐31‐ and IL‐31RA‐positive cells.AnimalsFour atopic dogs sensitised to house dust mites.Methods and materialsSkin and blood samples were obtained 0 h, 24 h, 48 and 96 h after allergen provocation. IL‐31 and IL‐31RA single‐staining immunofluorescence (IF), as well as IL‐31/CD3, IL‐31/CD4 and IL‐31RA/β3‐tubulin double‐staining IF were performed. The IL‐31‐positive cells were counted subjectively.ResultsThe peak IL‐31 expression for three of four dogs occurred 24 h or 48 h postchallenge; it started to decrease at 96 h. There was no significant correlation between the IL‐31 expression scores and the serum IL‐31 concentrations or the macroscopic skin lesion scores (P = 0.35 and P = 0.36, respectively). The majority of IL‐31‐positive cells were positive for CD3 (range 91–100%) and CD4 (range 63–100%), indicating that they were helper T (Th) cells. Unexpectedly, sebaceous glands were strongly immunolabelled with IL‐31; the extinction of this positivity after immunoabsorption with IL‐31 further supported the validity of this immunostaining. The IL‐31RA was visualised on keratinocytes and a small proportion of dermal nerves.Conclusions and clinical importanceThe early and transient production of IL‐31 by Th cells supports the concept of using IL‐31 inhibiting strategies as a proactive therapy to prevent flares of AD skin lesions.