Self-reported subjective judgments about well-being, mood, or mental state are at the core of analytical and empirical tools in many social sciences. However, technological advances (particularly in neuroscience) are opening new ways of monitoring physiological processes through non-intrusive means. Such dense continuous data provide new and fruitful avenues for complementing self-reported data with a better understanding of human dynamics and human interactions. Thus, in this study, we follow more than 300 individuals over a period of 24 hours, mapping their core activities (5,000 recorded activities in total) with measurements of their heart rate variability (HRV) and their assessment of subjective hedonic well-being (positive affect). Our results indicate a robust correlation between physiological measurements and self-reported affect. A lower HRV is positively correlated with self-reported positive affect. We also find that physical activities as well as relaxation are important for well-being, no matter what sort of actions follow, increasing positive feelings and inducing positive spillover effects on later endeavours. Individuals also report higher positive feelings prior to performing physical activities, providing new insights on how to harness activity data to extrapolate to the future.