N-2 fixation, nitrate uptake, and tissue N levels were compared in four novel self-fertile inbred lines (referred to as A, B, C and D) of white clover (Trifolium repens L.) believed to differ markedly in their nitrogen relations. Plants were inoculated with a mixture of Rhizobium strains and grown in flowing nutrient solutions without mineral N for 18 d prior to supplying half the plants with 20 mu M NO3- over a 36 d treatment period. Net uptake of NO3- was measured automatically on a daily basis. The remaining plants constituted 'controls' solely dependent upon N-2 fixation for acquisition of N. Total uptake of NO3- over 36 days was 27.4 (A), 23.8 (B), 13.8 (C) and 30.1 (D) mmol N/plant, compared with N-2 fixation of 7.2 (A), 16.9 (B), 26.2 (C) and 6.2 (D) mmol N/plant by the NO3--fed lines. Total N2 fixation by corresponding 'control' plants was 15.6 (A), 35.6 (B), 35.3 (C) and 2.2 (D) mmol N/plant. Line D plants exhibited a 3:1 segregation with respect to either 'nil' or 'partially impaired' N-2 fixation; the corresponding ratio for line A was 1:3. Line C plants had a 'low NO3- uptake' phenotype characterised by specific NO3- uptake rates approximately half those of the other lines, higher N-2 fixation in the presence of NO3-, and low leaf + petiole concentrations of NO3-, although concentrations of NO3- in stolen and root fractions were similar in all lines. These inbred lines provide a tool for (1) the analysis of the genetic and physiological basis of variation in the efficacy with which NO3- uptake down-regulates N-2 fixation, and (2) testing ecological hypotheses for temporal and spatial variation in clover/grass coexistence in pastures.