The present study evaluated the effect of Instructional Consultation (Rosenfield, 1995) on the academic achievement of third through fifth grade students. Students whom teachers did (n = 201) and did not (n = 8119) select as the focus of consultation were balanced on their estimated propensity to be selected using logistic regression of observed covariates. Multilevel modeling compared students in the two treatment conditions on teacher assigned grades and standardized measures of reading and math, net of prior achievement. A small, but statistically significant negative effect of the program (d = -.13) was found for standardized measures of math. No significant differences were found on the other outcome measures. Limitations include model misspecification, missing data, and treatment diffusion.