The growth analysis method was used to study the effect of 5 different sowing dates on the growth dynamics of 5 maize hybrids with different vegetation periods, sown in a small-plot field experiment in 1995 and 1996. The sowing dates were as follows: in 1995: Apr. 13, Apr. 20, Apr. 29, May 6, May 16; in 1996: Apr. 13, Apr. 23, May 3, May 16, May 24. The hybrids were: Mara (FAO 290), Norma (PAO 370), Maya (FAO 430), DK 524 (FAO 460) and Florencia (FAO 530). The soil of the experimental area was a humous loam of the chemozem type with forest residues, mildly acidic in the ploughed layer, moderately supplied with phosphorus and well. supplied with potassium. Under the experimental conditions there was a consistent decrease in both the effective heat sum (degrees C) and the photosynthetically active radiation (PAR) in the late sowing date treatments, compared with the early and optimum sowing dates. This reduction was especially great in the second half of reproductive growth. The growth dynamics of the reproductive mass, like that of total dry matter, could be described by a sigmoid curve. Two phases could be distinguished in the effect of sowing date on the growth dynamics of the reproductive mass. In the first phase the reproductive mass of the hybrids was smallest in the early sowing date treatment and greatest in the case of late sowing. In the second phase there was a substantial reduction in the reproductive mass of the maize hybrids in the late sowing date treatment. The sowing date influenced both the rate and duration of growth in the linear phase of grain filling. The duration of grain filling decreased considerably when sowing took place late. The grain yield was greatest in the optimum (Apr. 20 and Apr. 23) and early (Apr. 13) sowing date treatments, decreasing significantly when sowing was late (after May 10). Late sowing led to yield losses of 22.4% in 1995 and 16.1% in 1996. In late-sown crops there was a significant rise in the grain moisture content, from 14.4% to 19.6% in 1995 and from 22.9% to 27.8% in 1996. In both years the sowing date had a significant effect on the grain number per plant, while its influence on the thousand kernel mass varied with the year. In the reproductive growth phase the crop growth rate (CGR) had the highest average value in the early and optimum sowing date treatments (24.9-29.2 g.m(-2).day(-1)), gradually decreasing to 14.9-16.2 g.m(-2).day(-1) as the sowing date became later.