In near infrared (NIR) spectra, there could be observed the baseline drift, background noise, scattering effects, and overlapped peaks. These errors can disturb the robustness and reliability of multivariate calibration models. Influences of spectral pretreatment methods on calibration model were studied. The partial least squares (PLS) were applied for developing model of the precuring degree. The multiplicative scatter correction gave the best values for R-2 and RMSEC. R-2 was 0.96, RMSEC was 0.112, respectively. The method NIR and reference method were compared using Student's t test (alpha = 0.05) for paired values, the result showed that there was no significant difference between the NIR method and the reference method. (C) 2011 Wiley Periodicals, Inc. J Appl Polym Sci, 2012