Selective grazing by ruminant livestock and other herbivores causes difficulties in measuring or estimating the chemical composition and functional properties (digestibility and intake) of the diets of free grazers. New methodologies using faecal near infrared reflection spectroscopy (NIRS) offers scope to estimate diet quality in grazing animals once suitable calibration equations have been developed. This study was conducted to determine the potential of faecal NIRS to predict crude protein (CP), neutral detergent fibre (NDF), acid detergent fibre (ADF) and organic matter digestibility (OMD) of the diet of cattle grazing tropical grass pastures as well as organic matter intake (OMI). Reference data and faecal spectra were measured from a pen experiment in which 11 Creole steers (were individually housed and fed diets of fresh grass harvested from irrigated plots of (i) Digitaria decumbens and (ii) Dichanthium spp. The experiment ran for 70 days (14 days adaptation, 56 days measurement) and variation in diet quality was achieved by varying the age of grass regrowth after an initial mowing. Grass samples and faecal samples were bulked within each week of the measurement period within steer. The 88 dried and milled faecal samples (11 steers × 8 weeks) were scanned in a NIR Systems 5000 monochromator. Faecal spectra and reference data were used to calibrate and cross validate equations for predicting the various parameters using the Modified Partial Least Squares (MPLS) procedure. Derived standard errors of calibration (SEC) and coefficients of determination (R2) were 0.33 and 0.98% for dietary CP, 0.96% and 0.88 for NDF, 0.81% and 0.89 for ADF, 0.021 and 0.72 for OMD, and 4.62 g/kg of (body weight)0.75 and 0.61 for OMI. These values compared favourably with published faecal NIRS reports, where available for CP and OMD, and also with many published forage NIRS results and confirmed the potential of faecal NIRS as a technology for reliably predicting the chemical compo