As human populations encountered various selection pressures in different environments throughout history, local adaptation on traits affecting reproductive success occurred. Research over the last few decades have used genetic and phenotypic data to identify traits have been locally adapted in human populations and understand the genetic basis of adaptation. A majority of the established examples of adaptive traits in humans have been phenotypes with very few or singular genetic variants of large effect contributing to their phenotypic variation, but these cases are not necessarily representative of the spectrum of selection on traits. Many of the methods used to identify natural selection in humans screen for regions of the genome or genetic variants with the most extreme signatures of selection. As most complex traits have a polygenic basis, this screening strategy is not only limited in identifying adaptive traits in humans but also in characterizing the genetic basis of trait adaptation. As such, the human evolutionary genetics field has few good examples of selection on polygenic phenotypes. The gap in understanding selection on complex traits in humans has become more addressable with the growth of large genome-wide association studies (GWAS) data that uncover the genetic basis or architecture of phenotypic variation in populations. The research in this dissertation examines directional selection on two traits showing marked differentiation in human populations that represent relatively the low and high ends of the spectrum of polygenicity, skin pigmentation and height, respectively. Specifically, I examined light skin pigmentation evolution in Europeans and short stature evolution in the Baka of Central Africa. Skin pigmentation in Europeans lacks a typical signal of polygenic selection, suggesting an oligogenic adaptive response, whereas there is suggestive evidence for polygenic selection for short stature in the Baka. Finally, I examined the genetic architecture of height in sub-Saharan Africans broadly, and demonstrated with partitioned heritability analysis that even the most powerful GWAS fails to detect regions of the genome that harbor genetic variation that influences height in sub-Saharan African populations. This dissertation advances our knowledge of local adaptation in humans while also highlighting the current issues in testing for polygenic selection.