The conservation and the preservation of existing masonry buildings, most of them recognized as cultural heritage, require retrofitting techniques that should reduce the invasiveness and assure reversibility and compatibility with the substrate. In this perspective, the strengthening system should be able to improve the bearing capacity of the structural member and, at the same time, to assure mechanical and material compatibility. The use of Fabric Reinforced Cementitious Matrix (FRCM) composites is now recognized to be suitable for these purposes. Indeed, the inorganic matrix has comparable properties with respect to the existing historical mortars while the fabric has relevant tensile strength. At the same time, these systems assure satisfactory level of reversibility (or at least removability). In this scenario, the present research aims to investigate the FRCM-confinement of masonry columns focusing on the influence of specific parameters, still poorly investigated, in order to deeply understand their effect on the mechanical response. In particular, the experimental variables are: the type of masonry substrate (Tuff and clay brick with lime-based mortar), the type of FRCM system (glass dry mesh + lime-based mortar and steel mesh + lime-based mortar) and the number of plies (1, 2 and 3). In addition, a detailed experimental characterization of the utilized materials has been carried out, including bond tests between the reinforcement and the substrate. The results evidenced that the FRCM is an effective solution for masonry columns confinement once a proper design is performed, taking into account all involved parameters. The different strengthening systems exhibited different failure modes. Generally, a single ply of external reinforcement produced a negligible increase of bearing capacity. Both strengthening systems applied with multi-ply strengthening schemes produced a significant increase in terms of strength and ultimate axial deformation. This benefit was observed for both Tuff and clay masonry columns. Two available design-oriented formulas, reported in the Italian CNR (National Research Council) and ACI (American Concrete Institute) guidelines have been utilized, in order to further investigate their accuracy, mostly in the case of multi-layered reinforcement. The performed comparisons highlight that the two design relationships provide similar and accurate results when referred to the GFRCM (Glass-FRCM) system in 1- and 2-layer configurations, while the predictions appear conservative when 3 layers of GFRCM are utilized, for both masonry types. Considering the SRG (Steel Reinforced Grout) system, the results predicted by the two models are more scattered, mostly when the number of layers increases. In addition, the formulation proposed by CNR appears more accurate in the case of Tuff masonry while the ACI predictions are closer to the experimental results in the case of clay brick masonry.