The events we encounter and the emotions we experience are valenced-they are positively or negatively charged. Although these occurrences seem to be distributed irregularly throughout the day, the two experiments presented here reveal systematicity in the temporal dynamics of affective experience using a variety of time-series analyses. In Experiment 1, participants used a portable button to respond to event valence (the positive or negative charge of an event in the environment) or affective valence (one's positive or negative feeling at the time of responding). This methodology yields signed response durations, indexing the valence and intensity of an occurrence, and inter-response intervals, indexing their distribution. These measures revealed that valenced occurrences are correlated with both temporally proximal and remote occurrences. Experiment 2 validated the methodology employed in Experiment 1 using artificial, laboratory-created event structures. Implications of dynamical approaches to understanding emotion are discussed.