Quantitative proteomics plays an important role in validation of breast cancer related biomarkers. In this study, we systematically compared performance of label-free quantification (LFQ) and SILAC with shotgun and directed methods for quantifying breast cancer related markers in microdissected tissues. We show that LFQ leads to slightly higher coefficient of variation (CV) for protein quantification (median CV=16.3%) than SILAC quantification (median CV=13.7%) (P<0.0001), but LFQ method enables ~60% more protein quantification and is also more reproducible (~20% more proteins were quantified in all replicate samples). Furthermore, we describe a method to accurately quantify multiple proteins within one pathway, i.e. "focal adhesion pathway", in trace amounts of breast cancer tissues using a SILAC-based SRM assay. Using this SILAC-based SRM assay, we precisely quantified five "focal adhesion" proteins with good quantitative precision (CV range: 2.4%-5.9%) in replicate whole tissue lysate samples, and replicate microdissected samples (CV range: 5.8%-16.1%). Our results show that in microdissected breast cancer tissues, LFQ in combination with shotgun proteomics performed overall best and is therefore suitable for both biomarker discovery and validation in these types of specimens. The SILAC-based SRM method can be used for development of clinically relevant protein assays in tumor biopsies.