Seventeen experiments were conducted on farmers' fields to study the effect of native Zn on the fertilizer Zn needs of rice in an extensive soil series (Rarha; Udic Ustochrepts) of Kanpur, Uttar Pradesh, India. Rice (cv. Saket-4) was grown with four rates of fertilizer Zn (0, 1.25, 2.5 and 5 kg Zn ha-1) in the soils of varying DTPA-Zn status. Application of 5 kg Zn ha-1 resulted in significant increases in the dry matter yield (grain + straw), Zn concentration, and uptake of Zn by the crop in soils having DTPA-Zn less than 0.60 ppm; the magnitude of the response decreased with increasing DTPA-Zn status of the soil. Fertilizer Zn needs for optimum grain yield were 7.10 and 3.73 kg ha-1 in soils of low (<0.60 ppm) and medium (0.61-1.00 ppm) DTPA-Zn status. Recovery of fertilizer Zn was higher in low Zn soils compared to high Zn soils. DTPA-Zn status of the soil and Zn concentration in the third leaf of 45-day-old plants showed a highly significant correlation with percentage grain yield. The critical limits of DTPA-Zn and leaf-Zn for predicting response of rice to fertilizer Zn were 0.80 and 26.5 ppm, respectively.