Historians have generally presumed that Irish immigrants in the late nineteenth century suffered from ethnic job discrimination. However, empirical scholarship reports conflicting evidence. The present article presents new evidence on the issue based on data from Major League Baseball circa 1880. These data are unique in that “firms” (teams) and individual “employees” (players) can be identified along with “job assignments” (positions played) and “performance” (e.g., batting averages). Linking the players' names with U.S. census enumeration records allows relatively accurate identification of ethnicity. I test various hypotheses derived from Gary S. Becker's economic theory of discrimination. The main results are that Irish players outperformed non-Irish players both on average and at the margin, were (generally) relegated to less central positions in the field, were more often required to fill in at nonregular positions, and were less likely to be hired as managers. In addition, the proportion of Irish on ball clubs and in their host cities was positively correlated, and team win percentage had a (weak) positive correlation with the team's proportion of Irish. Overall, the results generally support anti-Irish discrimination against skilled workers in this highly visible, albeit small, “industry.”