Plant-to-plant variability has been observed to affect maize grain yield negatively. The objective of this study was to investigate the response of plant-to-plant variability to plant density with maize hybrids from different breeding eras in Ontario, Canada. In 1997 an old (Pride 5) and a new (Pioneer 3902) maize hybrid were grown at a low (3.5 plants m\sp−2) and a high (11 plants m\sp−2) plant density at Elora, Ontario. At each density seeds were sown either all on the same day to produce uniform stands (control), or on alternative sowing dates of three to produce non-uniform stands (treatment). At the low density uniformly sown and non-uniformly sown stands did not differ in grain yield and plant-to-plant variability for grain yield. At the high density non-uniform stands yielded less than uniform stands in both hybrids and grain yield of the new hybrid was 30% higher than the old hybrid in the uniform stand and 46% higher in the non-uniform stand. A second experiment was conducted at Cambridge, Ontario in 1996 and 1997 to explore the association between genetic improvement and plant-to-plant variability among six maize hybrids released from 1959 to 1995. Hybrids were grown at three plant densities (3.5, 7.0 and 11 plants m\sp−2) and two soil nitrogen levels. A negative correlation (R\sp2 = 0.73) was found between plant-to-plant variability for individual plant yield and yield per unit area at the medium and high densities.