The feasibility was investigated for online detection of soluble solids content (SSC) of “Yuluxiang” pear by visible-near infrared (visible-NIR) diffuse transmittance spectroscopy. 358 samples were divided into the calibration and prediction sets (269∶89) for developing calibration models and assessing their performance. By analyzing, the Vis-NIR transmission spectra of 'YuLuxiang' pears have three peaks at 625 nm, 725 nm and 800 nm and three troughs at 625 nm, 725 nm and 800 nm, respectively. Different preprocessing approaches were tested, it was found that the best approaches were the combination of first derivative (1D), smoothing and multiplicative scattering correction (MSC) preprocessing methods. The partial least square (PLS) regression and least square support vector machine (LS-SVM) models were developed with the pretreatment methods by the combination of first derivative (1D), smoothing and multiplicative scattering correction (MSC). The new samples of the prediction set were applied to evaluate the performance of the models. Compared with PLS model, the performance of LS-SVM model was better with the root mean square error of prediction (RMSEP) of 0.316% and the correlation coefficient of prediction of 0.949. And the spectral dimension reduction method of principal component analysis (PCA) and the kernel function of radial basis function (RBF) were suitable to improve the predictive ability of the LS-SVM model. The results suggested that it was feasible for online detection of SSC of ‘Yuluxiang’ pear by visible-NIR diffuse transmission spectroscopy combined with LS-SVM algorithm. The online detection of soluble solids content (SSC) of “Yuluxiang” pear by visible-near infrared (visible-NIR) diffuse transmittance spectroscopy was demonstrated.