The subjective well-being or happiness of individuals is an important metric for societies, but we know little about how the cumulative influence of daily life events are aggregated into subjective feelings. Using computational modeling, we show that momentary subjective well-being in a decision-making task is explained not by task earnings, but by the combined influence of recent reward expectations and prediction errors arising from those expectations. Using functional MRI we also find that neural responses in the striatum relate to changes in subjective feelings. Dopamine neurons are thought to represent reward prediction errors, signals that can be used for learning, and striatal neural responses may relate to dopaminergic input. By pharmacologically manipulating dopamine levels with the dopamine precursor L-DOPA, we find that L-DOPA affects both the decisions that subjects make and the emotional reactions to decision outcomes, supporting roles for dopamine distinct from its established role in learning.