This paper investigates the evolution of human capital in China for 31 provinces over the period of 1985–2016 from a club convergence perspective. Per capita human capital stocks, estimated using the Jorgenson–Fraumeni lifetime income approach, are for the first time examined within a non-linear latent factor framework that allows to model a wide range of transition dynamics for each province along the path to convergence. The study finds no overall convergence between provinces in China, however, the results strongly support the existence of multiple convergence clubs. While a small group of provinces are converging toward the highest levels of human capital, most of the other provinces are failing to catch up and form separate clusters that converge to lower equilibria. These regional patterns provide new evidence on the increasing human capital gap between Chinese provinces, posing a significant challenge to a more inclusive and harmonious human and economic development.