With 37 zoysia seed samples with different germination rates ranging from 58.5% to 92%, harvested in different years from 2009 to 2011 and from different locations of China, a model for determining germination rate of zoysia seeds was tried to be built by near infrared reflectance spectroscopy with quantitative partial least squares (QPLS). All the seeds samples were divided into two groups: calibration set (including 28 samples) and validation set (including 9 samples). The results showed that with the spectral range from 6 000 to 7 000 cm-1 and 6 main components, there was a better fitting between the predictive value and true value. Determination coefficients (R2) of calibration and validation sets are 90.73% and 91. 80%, the coefficients of correlation are 0.986 6 and 0.987 2, the standard errors are 9.80 and 9.47, and the average absolute errors are 7.64% and 6.98% respectively. Even with different calibration samples, the models have a high determination coefficient (R2 over building of NIR model for determining 90%), low standard errors (about 10.00) and low absolute errors (about 8.00%). The building of NIR model for determining germination rate of zoysia seeds could promote the application of high quality seeds in production.