<p>Sugarcane is an important crop for sugar production in the US. Sugarcane genotypes from different breeding programs from United States were evaluated in Texas for release in that environment. Mixed models indicated that there was no consistent advantage between genotypes from different breeding programs. To understand selection performance and genotype by environment interaction (G × E), the sucrose content, cane and sugar yield of 4–5 locations from 4 series (Series 2011 to 2014) of plantings each comprising 3–4 crops in the late-stage breeding program were used to construct Genotype plus Genotype by Environment (GGE) biplots. These biplots were utilized to identify genotype characteristics such as performance and stability, and location characteristics such as macroenvironments and yield discrimination. Only Series 11 had a significant location by genotype by crop interaction for sugar and cane yields, so it was analyzed by crop, while the other series were analyzed across crops. GGE biplots indicated that sugar yield mega environments were inconsistent by crop in Series 2011, but had two mega environments when averaged, while Series 12 and 13 had only one mega environment and Series 14 had three. The checks CP 72-1210 and CP 89-2143 had good stability and yield across series and crops indicating that they were good comparison cultivars. Estimated heritability of sugar yield indicated that replication had the lowest contribution to heritability while ratoon contributed the most to heritability. The results indicated that the Texas sugarcane variety development program has adequate locations for evaluation of cultivars and that G × E is not consistent enough for location specific selection at this stage in the program.</p>