Grain yield potential of Australian wheat crops is often limited because of inadequate water for crop growth and grain filling. Greater early vigour, defined here as the amount of leaf area produced early in the season, should improve the water-use efficiency and yield of wheat crops grown in Mediterranean-type climates such as occurs in southern Australia. In order to maximise selection efficiency for early vigour in breeding programs, the magnitude and form of genetic variation for early vigour and its components was investigated for 2 contrasting wheat populations. The first population comprised 28 Australian and overseas wheat varieties evaluated in a serial sowing study in Canberra. The second population contained 50 random F 2:4 and F 2:6 families derived from a convergent cross of elite CIMMYT wheat lines evaluated in Canberra, and in the field at Condobolin, New South Wales. For the first population, environmental effects on leaf breadth and length, and to a lesser extent, phyllochron interval, produced significant (P < 0.05) changes in leaf area. Large and significant (P < 0.05) differences were observed among Australian and overseas wheats for early vigour and its components. Australian varieties were among the least vigorous of the lines tested, with a number of overseas varieties producing about 75% greater leaf area than representative Australian wheats. Increased leaf area was genetically correlated with increases in leaf breadth and length, and a longer phyllochron interval. Significant (P < 0.05) genotype ´ environment interaction reduced broad-sense heritability (%) for early vigour (H ± s.e., 87 ± 26) compared with leaf breadth (96 ± 25) and length (97 ± 27). Narrow-sense heritability (%) in the second population was small for leaf area (h2 ± s.e., 30 ± 6) and plant biomass (35 ± 7), but high for leaf breadth (76 ± 14) and length (67 ± 16). Genetic correlations were strong and positive for leaf area with plant biomass, leaf breadth and length, specific leaf area and coleoptile tiller frequency, whereas faster leaf and primary tiller production were negatively correlated with leaf area. The high heritability for leaf breadth coupled with its strong genetic correlation with leaf area (rg = 0.56-0.57) indicated that selection for leaf breadth should produce genetic gain in leaf area similar to selection for leaf area per se. However, the ease with which leaf breadth can be measured indicates that selection for this character either by itself, or in combination with coleoptile tiller production, should provide a rapid and non-destructive screening for early vigour in segregating wheat populations. The availability of genetic variation for early vigour and correlated traits should enable direct or indirect selection for greater leaf area in segregating wheat populations.