The linear growth rate is an essential parameter to describe and simulate the crystal growth processes of solid materials. In the present study, two independent methods were used to estimate how calcite particle size was increasing with reaction time. First, a direct method by using Rietveld refinements of X-ray diffraction patterns quantifies the variation of the coherent domain average size r with reaction time t. Second, we used a mass balance method where the formed calcite amount Mol(CaCO3.t) was determined as a function of time and the linear growth rate was deduced from mass growth rate. For both methods, a kinetic pseudo-second-order expression was successfully used to fit the data and estimate the initial growth rates of nanosized calcite. The results deduced from Rietveld refinements showed that such rates were roughly equivalent for two different temperature conditions, i.e. 0.35 nm/s at 30 degrees C and 20 bar, and 0.27 nm/s at 60 degrees C and 20 bar. However, these results were significantly different from those deduced from mass growth rate. For this case, values of 0.09 nm/s at 30 degrees C and 20 bar, and 0.06 nm/s at 60 degrees C and 20 bar were determined. This significant discrepancy could be ascribed to other simultaneous processes during crystal growth of calcite such as agglomeration and possibly dissolution-reprecipitation reactions, complicating considerably the measurement of linear growth rate by mass balance. The latter process was possibly enhanced by the release of water into the reactor during the gas-solid carbonation reaction (Ca(OH)(2) + CO2 -> CaCO3 + H2O), and suggested to be experimentally evidenced for reaction extents greater than 80%. Taken together, these results suggest that the method based on Rietveld refinements may be more reliable to determine initial linear growth rates for reactions initiated in biphasic (gas-solid) systems, whereas both methods were previously demonstrated to be equivalent for triphasic systems. (C) 2011 Elsevier B.V. All rights reserved.