UBC9 encodes a protein that conjugates small ubiquitin-related modifier (SUMO) to target proteins resulting in a change of their localization, activity or stability. Genetic variability may affect expression and activity of UBC9 and may have an impact on breast tumor progression. We investigated associations between UBC9 genotypes and histopathological parameters in 1,021 breast cancer cases of the GENICA collection using a single nucleotide polymorphism (SNP) tagging approach. Genotyping analyses were performed by TaqMan(R) allelic discrimination. Odds ratios (OR) and 95% confidence intervals (CI) were calculated by ordinal logistic regression. Multiple imputation based on HapMap data was applied to boost the power of the study. The study revealed significant associations of three UBC9 SNPs with histological grade (rs7187167, p(trend) = 0.001; rs11248866, p(trend) = 0.009; rs8052688, p(trend) = 0.008). Model selection identified a recessive penetrance model for rs7187167 as the best representation of tumor grade (global p = 0.001). This model did not improve by inclusion of additional SNPs in linkage disequilibrium. Imputation of SNPs in a 300 kb region around the genotyped SNPs supported rs7187167 as a major contributor to tumor grade. Compared with common allele carriers, rare homozygotes presented less frequently with high grade tumors (G3 vs. G1: OR 0.26, 95% CI 0.11-0.62; G3 vs. G2: OR 0.45, 95% CI 0.23-0.86). In addition to tumor size, nodal status and estrogen receptor status, multivariate analyses confirmed an independent role of rs7187167 as predictor of tumor grade (p = 0.0003). The present results underline the value of genetic variation in UBC9 for breast cancer prognosis.