Greenhouse experiments were conducted to determine the statistical precision of estimating herbicide dose-response treatment effects by covariance analysis (ANOCOVA) relative to standard analysis of variance (ANOVA). Analyses of corn seedling response to the translocated herbicides fluazifop-P, sethoxydim, and quizalofop at 10 to 60 g ai ha-1 indicated that treatment effects were estimated with 26 to 116% greater precision by ANOCOVA than ANOVA. Covariance analyses of treatment effects for corn response to the contact herbicides paraquat, acifluorfen, and lactofen at 50 to 300 g ai ha-1 gave 8 to 13% greater precision than ANOVA. Gains in precision by ANOCOVA for all experiments were generally greatest when shoot dry weight was analyzed as the response variable and pretreatment fifth leaf length served as the covariate.