Participants' estimates of a conversation's duration immediately after it occurs are based on their enjoyment of the conversation and were shorter when the person on whom they had focused talked a lot than when (s)he said very little. After a 2–3 day delay, how-ever, they base their estimates on the amount of the conversation they could remember and estimate that it lasted longer in the former condition than in the latter. These conclusions were confirmed in both simulated conversations and an actual interaction between partners of the opposite sex. Thus, individuals' immediate and delayed estimates of the duration of an interaction can be opposite in direction, and this difference is driven by both the characteristics of the conversation (i.e., speaker dominance) and social and motivational factors that could influence people's focus of attention (i.e., the attractiveness of their conversation partner).