I study a sequential entry game in a winner-take-all market where success depends on the relative quality of the project and where some entrepreneurs (but not all) overestimate the quality of their own project. For a potential early entrant, overestimating one’s project quality (i.e., individual optimism) can have both negative and positive consequences: it fosters her incentive to enter, which can lead to excess entry but can also limit underentry (in markets where realistic entrepreneurs stay out); it can also help deter rivals. I show that the net gain of individual optimism for entrepreneurs with low-quality projects (i) depends on market potential (the market’s capacity to provide positive payoffs for one or two entrepreneurial firms) and group-level optimism (the fraction of optimists vs. realists among entrepreneurs), (ii) is in general positive in low-potential markets where optimism helps deterring rivals with projects of higher quality, and (iii) is maximized at moderate levels of group-level optimism in low-potential markets and decreases with group-level optimism in high-potential markets. Finally, I show that overoptimistic beliefs are often persistent, particularly in low-potential markets, and that optimism (in comparison with realism) can facilitate learning.