Three issues are considered on variation in number of presacral vertebrae (PSV) in humans: (1) sexual difference in number of PSV, (2) inactivation of Hoxd-11 gene as etiology for a supernumerary lumbar vertebra, and (3) anatomical correlates of a supernumerary lumbar vertebra, including lumbar-sacral nearthrosis, and pelvic size. Sample was 407 skeletonized females and 1,318 males from United States; ages at death were 20 to 49 years. Two subsamples of males were used: (1) 98 with modal numbers of cervical, thoracic, lumbar, and sacral vertebrae (PSV = 24) and (2) 45 with a supernumerary lumbar vertebra but modal numbers for other vertebral segments (PSV = 25). Measurements were taken of ulna, second metacarpal, vertebrae, femur, and pelvis; presence of lumbar-sacral nearthrosis was observed. Although 90% of females and males have 24 PSV, females have higher frequency of 23 PSV and males have higher frequency of 25 PSV. Compared to males with 24 PSV, males with 25 PSV and supernumerary lumbar vertebra show (1) no difference in anatomies associated with inactivation of Hoxd-11, and (2) higher frequency of lumbar-sacral nearthrosis and smaller pelvic inlet circumference. Sexual difference in number of PSV may be due to tempo of somite formation and Hox gene activation. Hypothesis is not supported that a supernumerary lumbar vertebra is due to inactivation of Hoxd-11. The presence of a supernumerary lumbar vertebra is associated with small pelvic inlet circumference, which can be obstetrically disadvantageous.