We give a gap-free differentially private algorithm for the principal component analysis (PCA) problem with Gaussian data.