In this study, hyperspectral imaging technology was used to achieve rapid and nondestructive detection of key quality indicators of kiwifruits, including soluble solids content, pH, firmness, and color e value. A hyperspectral imaging system was employed to capture hyperspectral images of kiwifruits, and the average spectral reflectance values from regions of interest were extracted for each sample. Simultaneously, reference values for the quality indicators were measured according to Chinese national standard methods. The original spectra were preprocessed using four methods: Second derivative, first derivative, multi-scatter calibration, and standard normal variation (SNV). Based on the full spectra, partial least-squares regression and principal component regression models were developed to predict kiwifruit quality indicators, and the optimal spectral preprocessing method was selected. To enhance model efficiency and stability, characteristic wavelengths were selected from the full spectra using a successive projections algorithm and competitive adaptive reweighted sampling (CARS). Multiple linear regression (MLR) and error back-propagation neural network models were then constructed using these selected characteristic wavelengths. Comparison of established models using the characteristic wavelengths with those utilizing the full spectra revealed that the SNV-CARS-MLR model had the best ability to assess kiwifruit quality indicators. Specifically, the prediction correlation coefficients (R_p) for soluble solids content, pH, firmness, and color e value were 0.9552, 0.8870, 0.8799, and 0.9079, respectively. The respective root mean square errors of the predictions were 0.9088, 0.0715, 2.3746, and 0.0285, and the respective residual predictive deviation (RPD) values were 3.3800, 2.0349, 2.1350 and 2.2838. All RPD values exceeded two, indicating strong model performance. Finally, pseudo-colouring was employed to input pixel information from hyperspectral images into the SNV-CARS-MLR model to visualise the distribution and compare kiwifruits of different quality grades. In conclusion, hyperspectral imaging technology enables rapid, nondestructive detection of key quality indicators in kiwifruits.