In the years 2004-2006 a field experiment was conducted at the Maize Breeding Station in Kobierzyce (50o99’ N; 16o95’ E) on changeability of selected maize hybrid yields in relation to the dose and kind of herbicide applied. The experiment was established according to split plot method in four replications. The first factor in the experiment was herbicides Mustang 306 SE, Callisto 100 SC, Maister OD introduced in a basic dose and in a doubled one. Within the frames of each first factor treatment there was analyzed yield changeability of 8 maize hybrids considerably differing in earliness of maturing and origin: Gazelle, Sastra, Tripoli, Monumental, Claudia, PR39H32, Lober and Anju 248. Discriminatory and cluster analysis of maize hybrid yield changeability under the influence of selected herbicides revealed profound differentiation among the hybrids analyzed. Out of 6 differentiated environments, taking into account chemical protection, the hybrids Sastra and Monumental featured the highest yields, while Claudia and Tripoli were characterized by the lowest yields. Differentiated herbicide doses did not reduce yields of the examined maize hybrids, except for Mustang 306 SE applied in the dose of 1.2 dm3.ha-1 , which resulted in lower yield values than those following the variant recommended by the producer. The most considerable effect on maize hybrid yield changeability was recorded for herbicides Callisto 100 SC in 3.0 dm3.ha-1 dose, as well as Mustang 306 SE in the dose of 1.2 dm3.ha-1. The hybrid Claudia reacted in a different way to herbicide doses used as compared to the remaining treatments. The yields of maize hybrids PR39H32, AN JOU248, PR39H32 and Sastra proved to be very similar, which was confirmed by insignificant Mahalanobis distances between those treatments.

83-91