Midparent-offspring regression yields a 30 to 40% more precise estimate of heritability than parent-offspring regression (Lynch and Walsh, 1998). This study was conducted to estimate heritability of plant height, ear height, tassel branch number, and yield in maize [Zea mays L.]. Parents were grown in Ames, IA 2000. Two replications of the full-sib S0 families were grown in a 15 x 15 lattice at Ames, Ankeny, Crawfordsville, and Lewis, IA in 2001. Plant height, ear height, and tassel branch number were measured at Ames and Lewis. Yield was measured at all environments. The means of those families were regressed on the midparent value for each environment and averaged over environments. Ninety-five percent confidence intervals were constructed for each estimate. Heritability estimates for yield were consistent across environments with a combined estimate of 0.009 (-0.034, 0.052). Plant height and tassel branch number showed consistent estimates across environments with combined estimates of 0.419 (0.365, 0.474) and 0.340 (0.270, 0.410), respectively. Heritability estimates of ear height were 0.391 (0.334, 0.449) at Ames and 0.053 (-0.027, 0.134) at Lewis with a combined estimate of 0.079 (0.035, 0.155). Environmental correlation between parents and offspring may be upwardly biasing the heritability estimates at Ames. Selection would be effective for all traits measured. Gain would be realized quicker for plant height and tassel branch number than for yield and ear height.