Controlled environment studies have indicated that the reaction of hematoxylin with Al-stressed roots can be used as the basis for a rapid and reliable test for Al tolerance in perennial ryegrass. Preliminary screening experiments identified a high degree of variability in the intensity of hematoxylin stain found in the root apex of plants subjected to a uniform Al treatment (24 h; 100μM Al). In the more intensely stained roots, visual symptoms of Al toxicity were observed in those regions of the axial structure of the root where the stain was most pronounced. These symptoms (thickening of the root and distortion) were notably absent from the less intensely stained roots. In subsequent growth experiments, plants segregated according to root growth reactions in solutions containing 0 - 150μM Al into clearly defined Al-tolerant and Al-sensitive populations which corresponded with the initial assessments made in hematoxylin. Plants from different cultivars which were initially considered to show similar levels of Al tolerance (hematoxylin test) did not perform equally well in the growth experiments. These differences implied that plants which exhibit Al tolerance in hematoxylin do not necessarily have good root growth characteristics. This finding is discussed in the context of improving and updating procedures directed at identifying germplasm adapted to growing on acid, Al-toxic soils. The magnitude of the differences in root growth reactions in the presence of Al between Al-tolerant and Al-sensitive plants suggests that the genetic resources available in perennial ryegrass are sufficient to support a conventional breeding programme directed at improving the Al tolerance of this species.