This study investigated the role of soil chemical attributes in explaining yield variability in a commercial flooded rice system in southern Brazil during the 2024/2025 growing season. We hypothesized that spatial variability in rice yield is primarily controlled by soil fertility gradients and redox-induced accumulation of Fe and Mn under flooding conditions. The study was conducted in a 45-ha field, where 44 soil samples (0–20 cm) were analyzed for chemical attributes and irrigation water depth. Grain yield data were collected using a combine harvester equipped with a yield monitor and used to generate thematic and relative yield maps. Data were evaluated through descriptive statistics, Pearson correlation, and principal component analysis (PCA). Average grain yield was 10.35 t ha⁻¹, with a spatial variability range of 4.5 t ha⁻¹. PCA explained 76.4% of the total variance, with the first component representing a fertility gradient characterized by higher Ca, Mg, organic matter, and Si contents positively associated with yield, whereas Fe and Mn showed negative associations. High-yield zones exhibited greater cation saturation and lower Fe and Mn concentrations. These findings demonstrate that yield heterogeneity is driven by the interaction between soil fertility and redox-mediated micronutrient dynamics, highlighting the importance of site-specific nutrient and water management.