Purpose: Dog owners represent 40% of the population, a promising audience to increase population levels of physical activity. The purpose of this study was to develop and test the psychometric properties of a new instrument to assess social-cognitive theory constructs related to dog walking. Method: Dog owners ("N" = 431) completed the Dogs and WalkinG Survey (DAWGS). Survey items assessed dog-walking behaviors and self-efficacy, social support, outcome expectations, and outcome expectancies for dog walking. Test-retest reliability was assessed among 252 (58%) survey respondents who completed the survey twice. Factorial validity and factorial invariance by age and walking level were tested using confirmatory factor analysis. Results: DAWGS items demonstrated moderate test - retest reliability (? = 0.39 - 0.79; k = 0.41 -- 0.89). Acceptable model fit was found for all subscales. All subscales were invariant by age and walking level, except self-efficacy, which showed mixed evidence of invariance. Conclusions: The DAWGS is a psychometrically sound instrument for examining individual and interpersonal correlates of dog walking.