This study provides novel evidence that expert economic agents’ work-related activities are systematically influenced by the time of day. We use archival data derived from time-stamped quarterly earnings conference calls together with linguistic algorithms to measure and track the moods of executives and analysts at different times of the day. The evidence indicates that the tone of conference call discussions deteriorates markedly over the course of the trading day, with both analysts’ and executives’ moods becoming more negative as the day wears on. Capital market pricing tests reveal that the time-of-day-induced negative tone leads to temporary stock mispricings. Our findings are relevant because the diurnal variations in behavior documented in the context of quarterly earnings calls are likely to extend across other important corporate communication, decision making, and performance situations, leading to potentially significant economic consequences.