Potassium (K) deficiency symptoms of rice (Oryza sativa L.) are difficult to visually diagnose during reproductive growth. Critical tissue–K concentrations may change across time and within cultivars. Our goal was to evaluate rice yield response to K fertilization and define continuous critical Y‐leaf–K concentrations (YLKC) for the production of ≥95% relative yield in pure‐line and hybrid rice. Ten Y‐leaves (uppermost leaf with a fully developed collar) were collected weekly during rice reproductive growth from selected fertilizer‐K rates (0–150 kg K ha–1) in 13 trials with variable soil‐test K (32–164 mg kg–1). Trials were seeded either with a pure‐line (8) or hybrid (5) cultivar. Significant yield increases from K fertilization occurred in 38% of the trials. Pure‐line cultivars were responsive to K fertilization and rice without fertilizer‐K produced 66–90% of maximum yield produced by fertilized treatments (8,506–11,153 kg ha–1). Hybrid rice receiving no fertilizer‐K produced 96–99% of maximum yield (9,067–12,572 kg ha–1) regardless of soil‐test K. The YLKC increased with increasing fertilizer‐K rate and declined across time for K‐sufficient rice or remained relatively constant across time for rice marginally sufficient or deficient in K. Rice YLKC above 16.0 g K kg–1 between R1 and R2 maximizes pure‐line cultivars yield production. The critical YLKC declined to about 13.0 g K kg–1 between R2 and R3 but was less accurate for diagnosing K deficiency than samples collected before R2. The YLKC can be used to assess pure‐line cultivars’ K nutritional status between the R1 and R2 stages.