Companion animals have been identified as a unique source of social support and as contributors to mental wellbeing. This study uses the Experience Sampling Method to test whether this effect is due to stress-buffering. A total of 159 dog and cat owners responded to a series of randomly scheduled questionnaires on their smartphones. At each measurement moment, they reported in whether a pet is present at that moment and to what extent they have interacted with the pet. They also reported on stressful activities and events and on their current positive (PA) and negative (NA) affect. Multilevel regression analyses showed that when a companion animal was present (vs. absent) the negative association between stress and PA is less pronounced (event stress: B = 0.13, p = 0.002, 95% CI = 0.05; 0.21 activity stress: B = 0.08, p < 0.001, 95% CI = 0.04; 0.12). No additional main effect was revealed when tested in a subsample of records that reported low or no stress. Main effects were found for the presence of a companion animal on negative affect (B = 0.08, p < 0.001; 95% CI = 0.12; 0.05) and for interacting with a companion animal on positive affect (B = 0.06, p < 0.001; 95% CI = 0.04; 0.08). This shows that the presence of a companion animal buffers against the negative consequences of stress on positive affect, indicating stress-buffering as a mechanism behind the pet-effect. It is, however, not the only mechanism and more research is required to further elucidate how companion animals contribute to human wellbeing.