We recently reported on the development of a 4-protein-based classifier (PDCD4, CGN, G3BP2 and OCIAD1) capable of predicting outcome to tamoxifen treatment in recurrent, estrogen receptor positive breast cancer, based on high resolution MS data. In order to move this findings towards a clinical setting, a precise and high throughput assay to measure these proteins in a multiplexed, targeted fashion would be favorable to measure large numbers of patient samples. By coupling immuno-precipitation to multiple reaction monitoring (MRM) MS and stable isotope dilution, we developed a high precision assay to measure the 4-protein signature in 38 primary breast cancer whole tissue lysates (WTL). Furthermore, we evaluated the presence and patient stratification capabilities of our signature in an independent set of 24 matched (pre- and post-therapy) sera. We compared the performance of iMRM with direct MRM in the absence of fractionation and shotgun proteomics in combination with label-free quantification (LFQ) on both WTL and laser capture microdissected (LCM) tissues. Measurement of the 4-proteins by iMRM not only showed higher accuracy in measuring proteotypic peptides (Spearman r: 0.74 to 0.93) when compared to MRM (Spearman r: 0.0 to 0.76), but also significantly discriminated patient groups based on treatment outcome (hazard ratio [HR]: 10.96; 95% confidence interval [CI]: 4.33 to 27.76; Log-rank P < 0.001) when compared to LCM (HR: 2.85; 95% CI: 1.24 to 6.54; Log-rank P = 0.013) and WTL (HR: 1.16; 95% CI: 0.57 to 2.33; Log-rank P = 0.680) LFQ-based predictors. Serum sample analysis by iMRM confirmed detection of the 4 proteins in these samples. We hereby report that iMRM outperformed regular MRM, confirmed our previous high resolution MS results in tumor tissues and have shown that the 4-protein signature is measureable in serum samples.