Selecting high yielding genotypes with stable performance is the breeders’ priority but is constrained by genotype × environment (G×E) interaction. We investigated canola yield of 35 genotypes and its stability in multiple environment trials (MET) in south-western Australia and the possibility to breed broadly-adapted high yielding genotypes. The Finlay–Wilkinson (F–W) regression and factor analytic (FA) model were used to investigate the G×E interaction, yield and genotype stability and adaptability. The cross-over response in the F–W regression, substantial genetic variance heterogeneity, and the genetic correlations in the FA model demonstrated substantial G×E interaction for yield. Cluster analysis suggests low, medium and high rainfall mega-environments. F–W regression indicated that genotypes with high stability (e.g. low regression slope values) produced relatively low yield and vice versa, but also identified broadly adapted genotypes with high intercepts and steep regression slopes. The FA model provided a more detailed analysis of performance, dividing genotypes by positive, flat or negative responses to environment. In general, early flowering genotypes responded negatively to favourable environments and vice versa for late flowering genotypes. More importantly, a few early flowering hybrids with long flowering phases were consistently productive in both low and high yielding environments, showing broad adaptability. These productive hybrids were consistent with those identified earlier by high F–W intercept and slope values. Hybrids were higher yielding and more stable than open-pollinated canola, as was Roundup-Ready® canola compared to the three other herbicide tolerance groups (Clearfield®, Triazine tolerant, conventional). We conclude that yield stability and high yield are not mutually exclusive and that breeding for broadly adapted high yielding canola is possible. [ABSTRACT FROM AUTHOR]