Saline concentration and corresponding indoor hyperspectral reflectance spectra for seashore saline soil samples were taken as research objects.Characteristics of reflectance spectra and the distribution pattern of soil salinity were analyzed in detail.Accordingly

correlations between soil salinity and its reflectance spectra and transformation of spectra were discussed deeply. Then

estimation models of soil saline concen-tration were established by using multivariate stepwise linear regression and curve estimation method. Re-sults showed that the reflectance spectra ranging from 400to 900nm presented a feature of decreasing firstly and then gradual increasing

while from 600to 625nm

the reflectance spectra had a large difference for dif-ferent soil samples.There existed significant correlation between soil salinity and first derivatives of reflec-tance spectra and its inverse. The largest correlation coefficient was up to 0.739. Soil salinity can be best es-timated by using S function and the R2 was 0.555. The research may provide a new measure for further dis-cussing the rapid detection of soil salinity at a large scale.