There is a set of hypotheses commonly used in the literature to explain how pets affect human well-being (e.g., pets as social catalysts). Many studies are reported as giving results consistent with one or more of these, but they may not appear to be sufficient to explain the impacts of several pet-related activities on owner well-being. Confirmation bias may also overshadow the consideration of alternative mechanisms. This report aims to review and evaluate a range of psychosocial hypotheses that might help to explain how pets affect the well-being of their owners. This included a theoretical testing of the hypotheses against the recurring themes which emerged from four previously published qualitative frameworks relating to pet-related activities and their well-being outcomes. Twelve psychosocial hypotheses were generated and evaluated using this process: (1) social catalyst-repellent, (2) emotional contagion and empathy, (3) social support, (4) biophilia, (5) attributed fault, (6) social norms, (7) annoyance by noises, (8) routine, (9) caring, (10) exercise, (11) learning, and (12) affective touch. Only three presented potentially contradictory evidence (i.e., social catalyst-repellent, routine, and caring hypotheses), but closer examination revealed that these could not be rejected. These twelve hypotheses are a source of reference for a broader consideration of how pets might affect human well-being. Researchers are encouraged to use, test and/or challenge these hypotheses using established methods of scientific falsification in order to identify which are of the most important in relation to specific owner well-being outcomes.