We present the first theoretical convergence rates for constrained expected improvement, one of the most popular constrained Bayesian optimization methods.