Tissue testing was studied in field experiments between 1979 and 1985 to predict when sulfur ( S) fertiliser was required for pastures in high rainfall (> 650 mm annual average) areas of south-western Australia. The pastures comprised about half subterranean clover and annual ryegrass ( Lolium rigidum Gaud.), the major pasture species in the region. Tissue testing was done for each species, using: (i) whole shoots, the present method used by commercial laboratories in Western Australia; (ii) youngest open leaves ( legumes, YOLs) or youngest expanded blades ( grass, YEBs); (iii) old leaves and blades ( leaves that were not YOLs or YEBs); and (iv) stems ( left after removal of YOLs, YEBs, old leaves and blades). Dried tissue was measured for total S, sulfate S, the total nitrogen: total S ratio and the sulfate S: total S ratio. For each pasture species, tissue test values were related to yield of dried herbage of that species measured for plants cut at ground level. Fertiliser nitrogen was not applied in the experiments.Annual ryegrass showed no significant yield responses to applied fertiliser S for all harvests in all experiments. Subterranean clover showed significant yield responses for most harvests of all experiments. At each site in each year, yield responses to applied S tended to become larger as the growing season progressed. For subterranean clover critical S values related to 90% of the maximum ( relative) yield varied for different harvests of the same experiment within and between years, and for different experiments in the same and different years. As determined from all data, critical S values were similar for all plant parts ( whole shoots, YOLs, old tissue, stems), with no consistent, systematic trend with plant age, and were: total S, range 0.10 - 0.30% S, mean 0.23%; sulfate S, range 0.01 - 0.14%, mean 0.04%; total N: total S ratio, range 11 - 30, mean 19; sulfate S: total S ratio, range 0.01 - 0.48, mean 0.27. The exception was that total S was lower for clover stems, the range being 0.06 - 0.20%, and mean 0.13%. Therefore, % total S in clover shoots can continue to be used as an indicator of sulfur deficiency in subterranean clover in the region.