Two field experiments were conducted at Southedge Tobacco Research Station, Mareeba, Queensland, Australia, in 1986 and 1987 to determine the rate of N and optimum fertilizer mixture with respect to P, K, Ca, and Mg to produce quality flue-cured tobacco (Nicotiana tabacum L. cv. ZZ100) on a granitic sand (Typic Paleustalf). Four rates of an experimental fertilizer mixture was evaluated at 4 rates of N (72, 96, 120, and 144 kg/ha). The mixture consisted of P, K, S, Ca, and Mg set at 100, 75, 50, and 25% relative to N, respectively, with the 100% mixture being 7N:2P:14K:2.5S:4Ca:1Mg. The effect of treatments on (a) plant nutrient accumulation, (b) chemical constituents of the cured leaf, and (c) extractable nutrients in the 0-30 cm soil depth was determined. The optimum fertilizer mixture which maintained agronomic performance of tobacco was 120 kg N/ha at 25% (7:0.5.3.5:0.6:1:0.25; N:P:K:S:Ca:Mg, respectively). Varying the fertilizer mixture from 100% to 25% of the macroelements relative to N had no significant effect on the ten agronomic parameters measured. As N rate increased, consistent significant differences occurred in both experiments for green grades, leaf price, and nutrient accumulation by the plant: for N in root, stem, leaf, and whole plant: for K, Ca, and Mg in leaf and whole plant; and for Mg in root and stem for one experiment. The cured leaf concentrations of N, K, Ca, reducing sugars, and reducing sugar:alkaloid ratio were significantly altered by increased N rate in both experiments, and only in one experiment for Mg and alkaloids. Adjustment of the macroelement ratio to 75, 50, or 25% significantly altered leaf levels of P, K, and Ca in both experiments and Mg in one. Soil analysis indicated that P and exchangeable K would decrease if the 50 and 25% fertilizer mixtures were used on a granitic sand for more than 2 years at 120 to 144 kg N/ha.