We evaluated 750 consecutive invasive breast carcinomas for HER-2/neu utilizing a combination of immunohistochemical (IHC) and fluorescence in situ hybridization (FISH) methodologies. IHC reactions of 3+ were considered HER-2/neu positive and 0 and 1+ IHC reactions were considered HER-2/neu negative. IHC reactions of 2+ were considered inconclusive and reflexed to FISH analysis. In addition, a 10% sampling and validation FISH analysis was performed on the positive and negative IHC tests. One hundred thirty-eight cases (18.4%) were HER-2/neu positive by IHC and/or FISH. One hundred twenty-three of the positive cases (89%) were 3+ IHC reactions and 14 positive cases were inconclusive by IHC and amplified by FISH. There was concordance with FISH in 77 of 78 (98.7%) of the positive or negative IHC cases that were tested (95% confidence interval [CI] = 93.1 to 100%). A single IHC-negative case showed HER-2/neu amplification by FISH. Thirty-nine cases were 2+ IHC (5.2%); 14 (36%) were amplified, 24 (62%) were not amplified, and one was not interpretable. HER-2/neu positivity was observed in 34% of grade 3 ductal carcinomas, 11.4% of grade 2 ductal carcinomas, 3.2% of grade 1 ductal carcinomas, and 3.2% of lobular carcinomas. Occasional cases with discordant IHC expression of HER-2/neu within the in situ and invasive carcinoma elements were also identified. IHC reliably characterized HER-2/neu in approximately 95% of the cases studied (95% CI = 93.0 to 96.2%) and was effective as a primary method for evaluating HER-2/neu status. In this study, 2+ IHC reactions were a heterogeneous group best regarded as indeterminate or inconclusive; in this series, only 36% were amplified by FISH analysis. Our findings suggest that a combination of IHC and FISH testing with FISH analysis performed reflexly on all 2+ IHC cases can optimize HER-2/neu testing.