Abstract: In order to study the influence of soil particle size on the estimate of soil organic matter (SOM) content by hyperspectral spectroscopy and obtain the appropriate soil particle size and reduce the workload, soil samples were sieved by 10, 20, 60 and 100 meshes, then the author scanned soil samples with hyperspectral data. The soil hyperspectral data were mathematically transformed into spectral reflectance (R), first derivatives of reflectance (R’) and the logarithm of the inverse of the reflectance [Log (1/R)]. Partial Least Squares Regression (PLSR), Support Vector Machine (SVM) and PLSR-SVM were used to build SOM content estimate model. The results showed that soil particle size had an obvious effect on the soil spectral reflectance, soil spectral reflectance increased with the decrease of soil particle size. The precision of SOM estimate model built by PLSR-SVM was better than the model built by PLSR and SVM. When the soil particle size was less than 0.25 mm, it was not helpful to improving the precision of this model. The results could provide guidance for soil samples’ pretreatment when using hyperspectral spectroscopy to predict SOM content.