Near infrared (NIR) reflectance spectroscopy was used to analyse samples (n = 332) from different soils from Uruguay (South America) for organic carbon (OC), total nitrogen (N) and pH. One set (n = 200) of samples randomly selected was used to develop the NIR calibrations while the remaining (n = 132) samples were used as the validation set. The samples were scanned in a small circular cup in reflectance mode (400?2500 nm), using a Foss NIRSystems 6500 (Silver Spring, MD, USA). Modified partial least squares (MPLS) was used to produce the calibration models and cross-validation was used to avoid collinearity effects among variables. Three mathematical treatments and four scatter corrections were also applied. The calibration coefficient of determination (R2 CAL) and the standard error in cross-validation (SECV) were 0.94 (SECV:1.9) for OC; 0.91 (SECV:0.19) for total N in g kg?1 and 0.93 (SECV:0.18) for pH, respectively. The simple correlation coefficient of validation (rVAL) and the standard errors of prediction (SEP) were 0.74 and 5; 0.73 and 0.4; 0.84 and 0.28 for OC, total N and pH, respectively.