Objective:To explore the feasibility and predictive value of Transit-Time Flow Measurement(TTFM) processed by Fast Fourier Transform(FFT) in evaluating the patency of grafts after coronary artery bypass grafting(CABG).Methods:The TTFM waveforms recorded during the operation of 114 CABG patients in our hospital from January 2015 to February 2017 were processed by FFT, the patients were followed up with CTCA after operation to evaluate the predictive value for graft failure.Results:320 grafts were grafted with the patency rate of 80.3%(257/320). The patency rate of LIMA group was 89.4%(101/113), and SVG group was 75.4%(156/207). H0, H1, H0/H1, and P1 of all grafts, H1and P1 in LIMA group, and H0, H1, P1 in SVG group were significantly different( P<0.05). In logistic regression, decreasing of H0( OR=0.92, 95% CI: 0.90-0.95) and increasing of P1( OR=2.26, 95% CI: 1.64-3.10) in all graft groups increased the risk of graft failure. In the LIMA group, increasing of H1( OR=3.57, 95% CI: 1.79-7.12) and P1( OR=1.53, 95% CI: 1.01-2.33) increased the risk of graft failure. In the SVG group, decreasing of H0( OR=0.83, 95% CI: 0.77-0.89) and H1( OR=0.05, 95% CI: 0.02-0.14), increasing of P1( OR=9.53, 95% CI: 3.04-29.86) increased the risk of graft failure. The ROC curve showed that H0 and P1 had a moderate degree of predictive accuracy in all graft groups. H1 and H0/H1 had a moderate degree of predictive accuracy in LIMA group, and H0 and H1 had a high degree of prediction, P1 had a moderate degree of predictive accuracy in SVG group. Conclusion:TTFM waveform processed by FFT has predictive value for the patency rate of CABG.