A two-year experiment was carried out at The Agricultural Research Station
of Gemmeiza, Gharbia Governorate in the seasons of 2001 and 2002. The aims of the
experiment were to study the effect 0 f nitrogen levels (0, 40, 8 0, 1 20 a nd 1 60 Kg
N/fed) on grain yield/fed of two maize hybrids namely: Single Cross 10 (S.C 10) and
Three Way Cross 351 (T.W.C 351), a yellow hybrid, to determine the degree of
response of grain yield to N fertilizer as well as to estimate the economic optimum N
rate. Also to investigate the relationship between grain yield/plant and its oomponents
using some multivariate techniques namely: correlation, stepwise multiple linear
regression and path-coefficient. Grain yield/fed of S. C 10 surpassed that of T.w.C 351 in both seasons. The
highest grain yield/fed of the two tested cultivars was produced by suppl~ing 120 Kg
N/fed in the two seasons. Quadratic model was the best of the tested models for
describing the relationship between grain yield of the two maize hybrids to fertilizer. The economic optimum N rate ranged from 101 to 110 Kg N/fed. Grain yield
produced by adding optimum N rate ranged from 3332.76 Kg/fed to 4509.89 Kg/fed
and return raged from £.E 1953.34/fed to £.E 2687.23/fed. The results of multivariate analysis techniques indicated that ear length, ear
diameter, number of kernels/row and weight of 100 kernels were the most important
contributing variables in the total variability of grain yield/plant. These variables have
to be ranked the first in breeding program for improving grain yield of maize.