One of the major challenges in biomarker development is the collection of tumor tissue of adequate quality for analysis. A prospective clinical trial was initiated to collect tissues from triple negative breast cancers prior to and after neoadjuvant chemotherapy in order to study the mechanisms of chemoresistance. Sixty patients had pre-chemotherapy biopsies performed by either a surgeon or a radiologist, while those with residual tumor after chemotherapy had research-only biopsies and/or surgical samples collected in liquid nitrogen, RNA-later and formalin. We examined each core for tumor cellularity, stromal content, and necrosis after which, RNA and DNA extraction was performed. We found that biopsies collected with ultrasound guidance were more likely to contain tumor than those collected by the surgeon. Patient reluctance to undergo research-only biopsies after chemotherapy was not a problem. Pre-chemotherapy tumor biopsies frequently did not contain any tumor cells (15%) or did not have ≥50% tumor content (63%). Indeed, 50% of patients had at least 2 pre-chemotherapy core biopsies with <50% tumor content. After chemotherapy, 30% of biopsy or surgical samples in patients with incomplete response did not contain any tumor. Finally, RNA-later not only made histopathological assessment of tumor content difficult, but yielded less DNA than fresh snap frozen samples. We recommend that high-quality tissue procurement can be best accomplished if at least three image-guided core biopsies be obtained per sample, each of these cores be examined for tumor cellularity and that at least some of them be freshly snap frozen in liquid nitrogen.