The feasibility and reliability of personalized medicine for cancer treatment is largely hampered by inefficient and oversimplified evaluation of anti-cancer drug response. Here, we report the generation and characterization of three-dimensional tumor slice culture (3D-TSC) platform for rapid anti-cancer drug discovery. Histological analysis of biomarkers and RNA sequence analysis demonstrate that 3D-TSCs accurately preserve cell repertoire and immune components of its original tumor. Using FRET based biosensor sensor C3 (C3) as an apoptotic reporter, we performed time-lapse high-throughput 166 anti-cancer drug screening on five cancer cell lines, which shortlists high efficacy drugs to screen human and murine 3D-TSCs and nominate effective candidates within 7 days after the surgery, a necessary speed for precision drug discovery. Further application of human and murine 3D-TSCs for immune check point blockade assays was successfully achieved with antibodies to PD-1 and/or PD-L1. Finally, revealed by label-free multi-photon fluorescence imaging, 3D-TSCs exhibit lipofuscin autofluorescence features for the time-course monitoring of drug response and efficacy. Overall, our time-course evaluation techniques on 3D-TSC platform demonstrate strong potential for predicting the efficacy of anti-cancer drugs and providing efficient and rapid drug screening platform for personalized chemotherapy and immunotherapy.