In 14 264 consecutive newborn babies the results of screening and early treatment for congenital dislocation and/or dysplasia of the hip (CDH) were evaluated in a prospective follow-up study. Barlow's test was done and the family history was recorded. Babies who were positive (n = 140) were immediately given abduction treatment. If the screening result was doubtful (133) or if the test was negative but the family history positive (685) the child was radiographed at 5 months and abduction treatment was started if any form of CDH was seen. Among Barlow-negative children with no family history (13,306), two reference groups were selected--at age 5 months (596) and at age 2 years (4365). Dislocation was missed at screening in 0.02%. Dysplasia was seen at 5 months in 15% of Barlow-negative children with a positive family history and in 2-3% of the reference group children. Crude estimates of the lower limits of validity indices for the screening test showed that the test is efficient in the identification of dislocation but probably has no value for dysplasia. Of the babies in whom treatment was started immediately 17% had relapse dysplasia after withdrawal of therapy, 3% had avascular necrosis, and 78% were normal at 2 years. When treatment was started at 5 months we found no relapse dysplasia, only 1% avascular necrosis, and 53-63% success at 2 years. In children with dislocatable hips we propose a wait-and-see treatment strategy, with early ultrasonography or radiography at 5 months.