The mineral carbonation of five different ultramafic mining residues was studied experimentally, namely chrysotile (Black Lake mine [B1] and Asbestos mine [Asb]), nickel (Ni-Cu Dumont mine project [Du] and Raglan Ni-Cu-EGP mine [Rg1]) and diamond mine residues (Renard mine project [Rnd]). The CO2 uptake, gas volume and mining residue physical and chemical characteristics were monitored in a fixed-bed diffusion cell to determine their potential for direct ambient carbon dioxide capture and to identify specific parameters influencing their reactivity. The various samples exhibited different behavior when reacted with gaseous CO2. Brucite- and (fibrous serpentine) chrysotile-containing residues exhibited higher pore-water pH, carbonation rate and yield with brucite content broadly dictating the reaction progress. Massive serpentines such as lizardite/antigorite-containing residues were found to weakly dissolve which significantly reduced their carbonation rate. The Mg-bru/Mg-tot ratio (brucitic Mg-total Mg atom ratio) and the percentage of fibers were thus the key parameters controlling the direct carbonation of alkaline mining residues. Recourse to alkaline and alkaline-earth metal contents to assess mining residue carbonation capacity may lead to inaccurate estimates as the residue Mg/Si and Mg/Fe ratios were found to greatly control the carbonation capacity through silica gel coating (high Si content) or/and iron (III) hydroxide passivation (high Fe content). (C) 2014 Elsevier B.V. All rights reserved.