The vertebral column plays a central role in primate locomotion and positional behavior. Understanding its evolution, therefore, has the potential to clarify evolutionary processes that have occurred in the primate lineage as well as the specific behaviors of extinct primates. However, to understand primate vertebral anatomy, it is important to determine how much of this anatomy is heritable and how much develops as a response to environmental factors during life. We estimated heritability for vertebral counts as well as typical cervical, thoracic, and lumbar elements from 210 individuals from the pedigreed Cayo Santiago Macaca mulatta skeletal collection. We found moderate heritability of vertebral counts (h2 = 0.216–0.326), but with strong heritability of the type of variation (e.g., a tendency to meristic or homeotic change) in the vertebral count (h2 = 0.599), suggesting a possible explanation for high variability in vertebral numbers among the hominoids in particular. The moderate heritability of vertebral count also suggests that vertebral count is an unsuitable metric for estimating the ancestral state for some taxa. We found strong heritability in the morphology of cervical and upper lumbar zygapophyseal facets (h2 = 0.548–0.550) and the thoracic spinous processes (h2 = 0.609–0.761), including high heritability of the spinous process angle in the upper thoracic and upper lumbar elements (h2 = 0.649–0.752). We suggest these are related to maintaining stability in the cervical and lumbar regions, and reducing motion in the thoracic region, respectively. We propose that spinous processes may contain greater phylogenetic information, whereas transverse processes may contain greater information of function ‘in life’. We also found important size effects, suggesting that size is the most heritable component of overall form and largely responsible for intertrait differences. This suggests that it is inappropriate to indiscriminately remove size effects from morphological comparisons.