The Atlantic silverside, Menidia menidia, is a widely distributed marine species along the east coast of North America that has been shown to exhibit local adaptation, despite evidence of high gene flow. This species displays a very strong and spatially fine scale increase in vertebral number with latitude, consistent with Jordan's Rule. Spatial and temporal variability can be seen on a microgeographic scale, potentially due to differences in site-specific developmental temperature and/or mixing among nearby locales. Most of the vertebral number variation is genetic however, and such tight clinal patterns implicate natural selection as the cause but its adaptive significance is unclear. Laboratory experiments show vertebral number responds to artificial selection on size, with populations with the largest size classes removed showing a decrease in vertebral number and vice versa. Natural selection on vertebral number is also evident in the wild, with vertebral numbers higher in juvenile populations compared to the same populations as adults. High latitude populations are thought to have evolved a greater number of vertebrae to allow for increased body flexibility in colder, move viscous water, however empirical evidence is limited. To test this theory, I hypothesized that at high temperatures, southern Atlantic silverside populations would show significantly higher critical swimming speeds than northern populations, but the reverse would be true at lower temperatures. Swimming speed experiments were conducted on southern (South Carolina) and northern (Nova Scotia) populations reared in a common environment. Each population was tested at a range of larval sizes and experimental temperatures. Vertebral number was negatively correlated with swimming speed at 28øC but such correlations at lower temperatures were non-significant. Few studies have investigated the link between vertebral number and swimming ability experimentally. The results of my research suggest extreme fine-tuning of vertebral number to natural selection in the wild and provide evidence for potential agents of selection.