This study examined the effect of Reading Recovery®, an early intervention program for first-grade children, on the rates of referral and placement in special education in New York City. Study participants consisted of students who had a complete Reading Recovery (RR) program (RR-discontinued and RR-recommended) and a comparison group composed of children with literacy difficulties, who were initially performing at a slightly higher level than the Reading Recovery children. The results showed that children who received the Reading Recovery program were referred for testing and placed in special education at a statistically significant lower rate. Reading Recovery is associated with a 5% reduction in estimated referral rate and a 3% decrease in placement rate. Additionally, an examination of the classification labels of children placed in special education revealed that RR-discontinued children were less likely to be classified as learning disabled than those in both the comparison group and the RR-recommended group. © 2002 Wiley Periodicals, Inc. Psychol Schs 39: 635–646, 2002.