We develop a real-time full-field optical angiography method using principal component analysis (PCA). In our approach, an undersampled laser Doppler method is used to record the raw images. Considering the difference in the signal component contributions, PCA is used to separate the dynamic blood flow and static background signals. The principal advantage of the PCA method is that the choice of a high pixel number can aid in efficiently extracting the blood flow signal with finite frame raw images, which can greatly improve the temporal resolution. Our phantom experimental results validate our choice of the optimal frame number for reconstructing an angiographic image. A vascular occlusion test on a rabbit ear demonstrates that global and simultaneous hemodynamic processes of vessels can be monitored.