North Dakota Soybean Variety Trial Results for 2025 and Selection Guide

The 2025 soybean growing season in North Dakota was marked by a cool, wet spring that delayed planting and reduced early-season yield potential across much of the state. In several northern and central counties, early rains followed by cloudy, cool weather caused poor emergence, forcing some growers to replant in early June. Despite this rocky start, a number of fields showed strong resilience, and yields still reached mid-40s in bushels per acre by harvest. Weather variability throughout the season, including alternating wet and dry periods, produced highly uneven crop performance, with statewide ratings hovering around 58% “good” but a substantial share in fair condition. Disease and pest issues were generally limited; however, cool, wet soils prolonged iron-deficiency chlorosis in susceptible areas and raised concern for SDS where soybeans followed corn. By mid-November, harvest was nearly complete, and yields mostly ended up in the mid-30s bu/ac range statewide, with better outcomes in fields that received timely moisture and avoided early stress. The agronomic data presented in this publication are from replicated research plots using experimental designs that enable the use of statistical analysis. The least significant difference numbers beneath the columns in tables are derived from the statistical analyses. If the difference between two varieties exceeds the LSD value, it means that with 95% or 90% probability (0.05 or 0.10 level), the higher-yielding variety has a significant yield advantage. If the difference between two varieties is less than the LSD value, then the variety yields are considered similar. The abbreviation NS is used to indicate “no significant difference” for that trait among any of the varieties. The coefficient of variation is a measure of variability in the trial and is expressed as a percentage. Large CVs mean a large amount of variation that could not be attributed to differences in the varieties. Only compare values within the table and look for trends for the desired trait among different experimental sites and years.