Many studies have examined Americans' immigration attitudes. Yet prior research frequently confounds multiple questions, including which immigrants to admit and how many to admit. To isolate attitudes on the former question, we use a conjoint experiment that simultaneously tests the influence of nine immigrant attributes in generating support for admission. Drawing on a two-wave, population-based survey, we demonstrate that Americans view educated immigrants in high-status jobs favorably, whereas they view those who lack plans to work, entered without authorization, are Iraqi, or do not speak English unfavorably. Strikingly, Americans' preferences vary little with their own education, partisanship, labor market position, ethnocentrism, or other attributes. Beneath partisan divisions over immigration lies a broad consensus about who should be admitted to the country. The results are consistent with norms-based and sociotropic explanations of immigration attitudes. This consensus points to limits in both theories emphasizing economic and cultural threats, and sheds new light on an ongoing policy debate.