An examination of tables from the 1950 U. S. Census of Population, and of the basic Persons punch card, shows that a few of the cards were punched one column to the right of the proper position in at least some columns. The result is that numbers reported in certain rare categories—very young widowers and divorcés, and male Indians 10–14 or 20–24—were greatly exaggerated. These errors occurred in spite of a careful checking program, and illustrate the necessity for users to view data concerning rare categories with special caution.