Abstract: The hyperspectral imaging technology was used to establish a rapid and non-destructive prediction method of banana fruit maturity in this study, aiming to solve the problem of fruit grade decline caused by rapid color change and putrescibility of postharvest banana. The dynamic changes of soluble solids content（SSC）, moisture content and hardness of banana fruits at different storage temperatures（20 ℃, 25 ℃ and 30 ℃） were measured,and contrastively analyzed with the corresponding spectral data in the same period. The characteristic wavelengths were selected by Monte Carlo partial least squares（MCPLS）, and the multiple linear regression（MLR） prediction model was established. The results showed that the determination coefficients of the prediction model of SSC, moisture content and hardness of banana fruits were 0.858 4, 0.873 5 and 0.912 8, respectively, which realized the nondestructive evaluation of SSC, moisture content and hardness at different ripening stages. It is feasible to apply hyperspectral imaging technology to rapid non-destructive testing of banana fruits quality parameters at maturation stage.