Zinc-phosphorus interactions have been frequently studied using a diverse number of crop species, but attainment of reproducible Zn deficiencies, especially severe ones, has been hampered by the use of conventional hydroponic solutions wherein contaminating levels of Zn are often near-adequate for normal growth. We utilized novel, chelator-buffered nutrient solutions for precise imposition of Zn deficiencies. Tomato (Lycopersicon esculentum L. cv. Jackpot or Celebrity) seedlings were grown for 15 to 18 d in nutrient solutions containing 200, 600, or 1200/xM P, and 0 to 91/zM total Zn. Computed free Zn 2÷ activities were buffered at ~< 10 -l°3 M by inclusion of a 100-/zM excess (above the sum of the micronutrient metal concentrations) of the chelator DTPA. At total added Zn = 0, acute Zn deficiency resulted in zero growth after seedling transfer, and plant death prior to termination. Free Zn 2* activities ~< 10-1°6 M resulted in Zn deficiencies ranging from mild to severe, but activities <~ 10-~2 were required to cause hyperaccumulation of shoot P to potentially toxic levels. Despite severe Zn deficiency (i.e. ca. 20% of control growth), tissue Zn levels were usually much higher than the widely reported critical value of 20 mg kg-~, which may be an artifact of the selection of DTPA for buffering free Zn 2÷. Across Zn treatments, increasing solution P depressed growth slightly, especially in Celebrity, but corresponding increases in tissue P (indicative of enhanced P toxicity) or decreases in tissue Zn (P-induced Zn deficiency) were not observed. The depressive effect of P was also not explained by reductions in the water-soluble Zn fraction. Within 40 h, restoration of Zn supply did not ameliorate high leakage rates (as measured by K + efflux) of Zn-deficient roots. Similarly, transfer of Zn-sufficient plants to deficient solutions did not induce leakiness within 40 h. Foliar sprays of ZnSO 4 almost completely corrected both Zn deficiency and membrane leakiness of plants grown in low-Zn solutions. Hence, maintenance of root membrane integrity appears to depend on the overall Zn nutritional status of the plant, and not on the presence of certain free Zn 2÷ levels in the root apoplasm.