In Western Australia yields of wheat are highly variable from season-to-season and farmers are seeking genetic solutions to increase yield stability. The tiller inhibition (TIN) gene has been proposed as a trait to stabilise wheat yields across seasons by reducing the investment in wasteful late tillers that do not contribute to yield. In theory a crop with fewer tillers would produce additional yield in poor seasons, but the cost is a yield penalty in good seasons. It is possible that this yield penalty could be overcome by sowing a great density of plants. In this research we tested the yield of near isogenic wheat lines either with or without the TIN gene across 5 sites in Western Australia. Trials were also sown at a range of densities from 25 to 300 plants/m2. The five sites had markedly different seasonal conditions and yield potential ranged from 0.7 to 3.2 t/ha. At most sites there was no interaction between plant density and genotype for yield or head number. Lines with the TIN gene always produced less yield than their free tillering partner. At the same population the TIN genotypes consistently yielded less than their free tillering partners. Across sites, yields of the TIN lines were approximately 65% of the free tillering lines. To achieve the same yield as the free tillering lines the TIN genotypes required a far greater sowing density. The effects on seed size, seed number, and protein content will also be discussed. This research has questioned the utility of the TIN gene to stabilise yields in this environment. The yield penalty could not be overcome by increasing sowing density. Furthermore, if fewer tillers in poor seasons is the objective then reducing plant population will be a far more cost effective option than using a genotype with the TIN gene.