Förster resonance energy transfer (FRET) microscopy provides unique insight into the functionality of biological systems via imaging the spatiotemporal interactions and functional state of proteins. Distinguishing FRET signals from sub-diffraction regions requires super-resolution (SR) FRET imaging, yet is challenging to achieve from living cells. Here, we present an SR FRET method named SIM-FRET that combines SR structured illumination microscopy (SIM) imaging and acceptor sensitized emission FRET imaging for live-cell quantitative SR FRET imaging. Leveraging the robust co-localization prior of donor and accepter during FRET, we devised a mask filtering approach to mitigate the impact of SIM reconstruction artifacts on quantitative FRET analysis. Compared to wide-field FRET imaging, SIM-FRET provides nearly twofold spatial resolution enhancement of FRET imaging at sub-second timescales and maintains the advantages of quantitative FRET analysis in vivo. We validate the resolution enhancement and quantitative analysis fidelity of SIM-FRET signals in both simulated FRET models and live-cell FRET-standard construct samples. Our method reveals the intricate structure of FRET signals, which are commonly distorted in conventional wide-field FRET imaging.