For decades, the efficacy of interventions in clinical trials enrolling dogs with atopic dermatitis (AD) relied on heterogeneous evaluations of skin lesions and pruritus using unvalidated tools. Although some instruments for clinical signs were validated later, there was little impact on standardizing outcome measures resulting in difficulties in comparing treatment efficacy between trials and impeding meta-analyses. Participants in the Outcome Measures subcommittee of the International Committee of Allergic Diseases of Animals (ICADA) collaborated for two years to develop a core outcome set (COS) for canine AD, the COSCAD. This project involved several steps, constantly-re-assessed during online exchanges, to define the scope of this COS, to identify the relevant stakeholders, the domains to be evaluated, the instruments available for measuring agreed-upon domains and how to express outcome measures. This COSCAD’18 was designed principally for therapeutic—but not preventive or proactive—clinical trials enrolling dogs with chronic, nonseasonal (perennial), moderate-to-severe AD. Selected domains were skin lesions, pruritus manifestations and perception of treatment efficacy. Instruments to evaluate these domains were the CADESI4 or CADLI, the 10-point pruritus visual analog scale (PVAS10) and the Owner Global Assessment of Treatment Efficacy (OGATE), respectively. The COSCAD’18 has three outcome measures: the percentages of dogs with veterinarian-assessed skin lesions or owner-rated pruritus manifestation scores in the range of normal dogs or those with mild AD; the third is a good-to-excellent global assessment by the pet owners of their perception of treatment efficacy. Importantly, this COSCAD’18 is not meant to represent the sole—or primary—outcome measures evaluated in a trial; authors are always free to add any others, which they deem will best assess the efficacy of tested interventions. Benchmarks to define a threshold for treatment success were not set, as what constitutes a clinically-relevant therapeutic efficacy is expected to vary greatly depending upon interventions. This COSCAD’18 should help veterinarians and owners compare the benefits of treatments in future trials. This COS should also facilitate the combination of trial results in future systematic reviews, thereby producing more reliable summary estimates of treatment effects and enhancing evidence-based veterinary dermatology.