In an effort to determine whether intelligence and school readiness could be estimated from observin g behavior and if middle class Head Start teachers ranked highly those pupils whose behavior resembled middle class behavior, the behavior patterns of 36 Head Start pupils were compared with their teachers' ranking of perceived intelligence and school readiness and with behavior patterns of 32 middle class and 40 upper-middle class children, all approximately age 4. The children's intelligence and school readiness were measured by standardized tests. Behavior patterns were observed on two schedules: continuous and summary. The results showed that (1) intelligence and school readiness were not correctly predicted from different types of behavior patterns, (2) while there was no difference in behavior among those Head Start pupils with actual intelligence differences, the behavior of those perceived to be brighter closely resembled middle class behavior, and (3) behavior patterns of all Head Start pupils were similar, regardless of actual school readiness. The data were tabulated, and suggestions were made for a replica study. A manual for using the observation schedule is included. (JS)