Sunflower (Helianthus annuus L.) hybrids do not respond identically to agronomic and environmental variables. The present study was conducted to quantify the influence of soil water potential and plant population on oilseed sunflower hybrids ‘Dahlgren 704’, ‘USDA 894’, and ‘Interstate 907’. Desire to refine knowledge of yield relationships was one motivation for the study, but a deeper stimulus was the need for basic knowledge of the detailed nature of sunflower stands. Hybrids were planted at four populations at two North Dakota sites in 1980. One site had Maddock (sandy, mixed Udorthentic Haploboroll) and Gardena (coarse‐silty, mixed pachic Udic Haploboroll) soils. This location received irrigation treatments limiting soil water potentials to minimums of ‐35, ‐70, and ‐105 J kg‐−1. The second site had Bearden (fine‐silty, frigid Aeric Calciaquoll) soil and was not irrigated. Yield increased with soil water potential from 201 g m−2 at – 105 J kg−11 to 284 g m−2 at ‐35 J kg−1. Dryland yields were 200, 179, and 164 g m−21 for hybrids 894,704, and 907. Population did not influence yields. However, irrigated and nonirrigated stand loss percentages increased with population at 0.8 and 2.2% per plant per square metre, respectively. Knowledge of the threshold population for stand loss provides a rationale for selection of population in the absence of yield differences because significant stand loss indicates a possibly undesirable level of intraspecific competition. Guidelines for irrigation scheduling for sunflower by use of tensiometers are now on a sounder experimental basis. Hybrids similar under irrigation yielded differently under dryland conditions.