Redshift-space clustering distortions provide one of the most powerful probes to test the gravity theory on the largest cosmological scales. In this paper we perform a systematic validation study of the state-of-the-art statistical methods currently used to constrain the linear growth rate from redshift-space distortions in the galaxy two-point correlation function. The numerical pipelines are tested on mock halo catalogues extracted from large N-body simulations of the standard cosmological framework, in the redshift range 0.5 z 2. We consider both the monopole and quadrupole multipole moments of the redshift-space two-point correlation function, as well as the radial and transverse clustering wedges, in the comoving scale range 10 < r[h −1 Mpc] < 55. Moreover, we investigate the impact of redshift measurement errors, up to δz ∼ 0.5%, which introduce spurious clustering anisotropies. We quantify the systematic uncertainties on the growth rate and linear bias measurements due to the assumptions in the redshiftspace distortion model. Considering both the dispersion model and two widely-used models based on perturbation theory, that is the Scoccimarro (2004) model and the Taruya et al. (2010) model, we find that the linear growth rate is underestimated by about 5 − 10% at z < 1, while limiting the analysis at larger scales, r > 30 h −1 Mpc, the discrepancy is reduced below 5%. At higher redshifts, we find instead an overall good agreement between measurements and model predictions. The Taruya et al. (2010) model is the one which performs better, with growth rate uncertainties below about 3%. The effect of redshift errors is degenerate with the one of small-scale random motions, and can be marginalised over in the statistical analysis, not introducing any statistically significant bias in the linear growth constraints, especially at z ≥ 1.