Flow velocity measurement based on speckle in optical coherence tomography (OCT) images is developed. Similar to traditional laser speckle signal, time-varying fluctuations in OCT speckle intensity is related to the average velocity of scattering particles within the sampling volume. Fluctuated signal of the speckle intensity is obtained from the detected OCT signal by filtering and demodulation. Then the spectrum of the speckle is deduced through Fourier transform of the fluctuated signal, and quantitative measurement of flow velocity can be conducted according to the calculated ratio of high to low (HLR) spectrum. The method of flow measurement based on speckle intensity in OCT instead of phase information is introduced. Investigation on the relationship of the calculated HLRs with the flow velocities is experimentally conducted. Flowing particles in a tube in mimic of capillary is visualized by the proposed method.