A further critical test of Horrobin's hypothesis that malignancy in cells may be dependent on gamma-linolenic acid (GLA) deficiency, has revealed that GLA supplementation produces a highly significant reduction in the growth rate (up to 87%) of a cultured human hepatoma cell line, compared with the growth rate of untreated hepatoma cells. This supports our previous suggestion that this hypothesis requires urgent further investigation at all levels including trials in human cancer patients.