This paper examines tests for hot- and cold-hand effects in men’s professional golf, based on score relative to par over sets of 3, 6, 9, and 18 holes. Using controls for each player’s annual performance and the difficulty of a set of holes on a given day, I find no evidence for a hot-hand effect. In contrast, I find evidence supporting the existence of a cold-hand effect—that is, a poor performance on one set of holes leads to a worse performance on the next set of holes. Simulations demonstrate that the cold-hand effect is quite large.