Background Stereotactic needle biopsy remains the cornerstone for tissue diagnosis for tumors located in regions of the brain that are difficult to access through open surgery. Objective We perform a meta-analysis of the literature to examine the relation between number of samples taken during biopsy and diagnostic yield, morbidity and mortality. Methods We identified 2416 patients from 28 cohorts in studies published in PubMed database that studied stereotactic needle biopsies for tumor indications. Meta-analysis by proportions and meta-regression analyses were performed. Results On meta-analysis, the morbidity profile of the published needle biopsy studies clustered into three groups: studies that performed < 3 samples (n = 8), 3–6 samples (n = 13), and > 6 samples during biopsy (n = 7). Pooled estimates for biopsy related morbidity were 4.3%, 16.3%, and 17% for studies reporting < 3, 3–6, and > 6 biopsy samples, respectively. While these morbidity estimates significantly differed (p < 0.001), the diagnostic yields reported for studies performing < 3 biopsies, 3–6 samples, and > 6 samples were comparable. Pooled estimates of diagnostic yield for these three groups were 90.4%, 93.8%, and 88.1%, respectively. Mortality did not significantly differ between studies reporting differing number of samples taken during biopsy. Conclusions Our meta-analysis suggests that morbidity risk in needle biopsy is non-linearly associated with the number of samples taken. There was no association between the number of biopsies taken, and diagnostic yield or mortality.