Purpose <p>Stereotactic brain biopsies are not without risk, and non-diagnostic biopsies can delay the instigation of appropriate treatment. Institutions rarely report their negative (non-diagnostic) results; this reporting is essential for quality and improvement in neurosurgical departments.</p> Methods <p>We report a retrospective single-center decade-long cohort of non-diagnostic stereotactic biopsies. We report the indications for biopsy, histopathological findings, rates of re-biopsy, lesion characteristics and biopsy success metrics such as depth-do-diameter, biopsy accessibility metrics and true (TN) and false (FN) negative rates.</p> Results <p>Of 387 biopsies, 23 were non-diagnostic (5.94%). Among 196 robot-assisted procedures, the non-diagnostic rate was 5.05%, compared to 6.88% across 189 non-robot-assisted biopsies (<i>p</i> = 0.52). The anticipated diagnosis was tumor/metastasis in 65% of cases. 26% of patients were asymptomatic prior to biopsy. The target was subcortical in 61% of cases, with median volume 8.36cm³ and lesion depth 65.05&#xa0;mm. Deeper lesions were larger (<i>r</i> = 0.47, 95%CI:0.02–0.76, <i>p</i> = 0.036). Biopsy showed indeterminate mild gliosis in 60.9%, normal parenchyma in 34.8% and hemorrhage in one case. Two patients underwent re-biopsy. Post-biopsy, 52% were treated medically, 26% remained under observation, the remainder were palliated or underwent surgery. On review of prior and future clinical/radiological/histopathological data, 7 (30.4%) were TN, 14 (60.8%) were FN. No TN biopsies occurred at a lesion depth of less than 50&#xa0;mm.</p> Conclusions <p>The optimal method and equipment for stereotactic biopsies remains an area of personal and institutional preference. Internal review of these metrics can identify areas for improvement of service delivery in a procedure that is central to the diagnosis and treatment of patients with brain lesions.</p>