Objectives. Prostate cancer detection in biopsies increases with the number of sites and total tissue sampled. Its dependence on needle core fragment length is uncertain. Methods. We surveyed two consecutive series of sextant needle biopsies from two practices in 1998 to 2000: 251 patients from Pennsylvania (group P) and 1596 from Virginia (group V). We tabulated the gross needle core lengths per sextant site and classified the diagnoses as benign or into four nonbenign categories: high-grade prostatic intraepithelial neoplasia; atypical small acinar proliferation, suspicious; atypical small acinar proliferation, suspicious plus high-grade prostatic intraepithelial neoplasia; and cancer. Logistic regression analysis was used to correlate cancer or a nonbenign diagnosis with the total length (sum of six sites) and, after excluding the sites with more than one core, with the length per single core, and the anatomic site of origin (apex, mid-gland, base). Results. The mean total tissue length sampled was 108 ± 27 mm (range 30 to 275) in group P and 81 ± 22 mm (range 30 to 228) in group V. Sextant sites with a single core contained a mean of 12.8 ± 3.5 mm tissue, with a 3.6-fold variation among the middle 95%. Group V core lengths at the apex averaged 11.8 mm, shorter (P = 0.0001) than mid (13.3 mm) or base (12.7 mm). A predictive value of longer length for a nonbenign diagnosis was noted in four of six sextants (P <0.04), with trend strongest at the apex, for which detection was influenced by abnormal digital rectal examination (P = 0.02) or ultrasound (P = 0.04) findings. Conclusions. The length of single cores sampled by sextant biopsy can vary more than 3.6-fold and represents a quality assurance consideration. The effect of length on cancer or nonbenign detection was maximal at the prostatic apex where the cores were shortest.