Abstract. An experimental calibration of the shifts of three major Raman peaks of quartz with hydrostatic pressure and uniaxial differential stress is presented, and implications for their use in geobarometry based on Raman spectroscopy of quartz inclusions are discussed. The position of the 206 cm−1 peak depends only on hydrostatic pressure P, and its pressure dependence is recalibrated with a peak-fitting procedure that is more adequate for Raman barometry than previous calibrations. The position of the 128 and 464 cm−1 peaks depends on P and also on differential stress σ, which can be determined from the position of these two peaks knowing hydrostatic pressure from the position of the 206 cm−1 peak. The results obtained here are different from those inferred previously from first-principles calculations. The present calibration provides direct relationships between Raman shifts and stress, with a simple formulation of residual pressure and differential stress assuming uniaxial stress along the c axis of quartz inclusions. It is tested on data from experimental and natural inclusions. Residual pressures from the present calibration are similar within uncertainties to those obtained with previous experimental calibrations. Residual differential stresses obtained from the 128 and 464 cm−1 peaks are very sensitive to the precision of Raman measurements. Experimental inclusions yield residual pressures consistent with synthesis pressure. Differential stresses obtained on some experimental inclusions are sometimes incompatible, providing a criterion for identifying inclusions under complex stress conditions that are not appropriate for geobarometry. Recent data on natural inclusions show self-consistent differential stress, consistent with the assumption of major stress along the symmetry axis of the inclusion crystals. The average pressure values from the 128 and 464 cm−1 peaks are similar to the residual pressure from the 206 cm−1 peak that depends only on hydrostatic pressure. It can be used to obtain pressure when the 206 cm−1 peak position cannot be used due to interference with host mineral peaks. Using the 128 and 464 cm−1 peaks alone, or averaging either 128 and 206 or 206 and 464 cm−1 peaks, can induce systematic bias in the residual pressure determination. Applications of the present results to natural inclusions suggest that combined determination of residual pressure and differential stress may be used for both barometry and thermometry pending further calibration.