Although many people strongly believe in wellbeing benefits of having pets (the ‘pet effect’), research on the ‘pet effect’ in actuality is highly mixed. Surprisingly, little research has explored the nature of (i.e., specific components of) the human–pet relationship, as a way to explain the contradictory findings. One such component is human–pet play, with play outside of the human–pet relationship being important for wellbeing and social buffering. Thus, the current quantitative study explores whether greater perceived play with pets contributes to greater wellbeing (‘pet effect’) and reduces anxiety during a time of acute stress (social buffering). The study employs a regression design recruiting men and women residing in the UK during the COVID-19 pandemic who own a dog and/or cat, with perceived pet play as the key predictor variable, and outcome variables of wellbeing and COVID-19 anxiety. Data were gathered and analysed cross-sectionally on day one (N = 189), and longitudinally over five days (N = 105), using multiple regressions. Overall, perceived pet play did not predict wellbeing nor COVID-19 anxiety. As such, the current study indicates human–pet play does not contribute to the ‘pet effect’ nor social buffering, thus raising questions for future research regarding the exact purpose of play within the human–pet relationship.