Efficiency of passive bioreactors (PBRs) to treat mine drainage (acid – AMD and neutral – NMD), in northern climates, is evaluated using pilot-scale PBRs installed at the Raglan Mine (Quebec, Canada). Three reactors (PBR-AMD, PBR1-NMD and PBR2-NMD) were filled with reactive mixtures and operated (at hydraulic retention times of 2.5d and 1d, for AMD and NMD) for 48 days, 94 days and 44 days, respectively. The AMD quality was 27.6 mg/L Ni, 23.7 mg/L Fe, 2.4 mg/L Cu, and 3186 mg/L SO42−, at pH 3.6, while the NMD quality was 23.9 mg/L Ni, 0.06 mg/L Fe < 0.003 mg/L Cu, and 587 mg/L SO42−, at pH 7.2. Removal efficiencies of the PBR-AMD reactor were 93–95 % vs 53–56 % Ni (at the beginning and end of the tests, respectively), 96–99 % Fe and 99 % Cu. In the PBR1-NMD reactor, 99 % vs 83 % Ni (in 2017 and 2018, respectively) was removed, while in the PBR2-NMD reactor, 95–99 % vs 61–83 % Ni (at the beginning vs at the end, respectively) was removed. However, all bioreactors failed to meet environmental criteria. None of the bioreactors developed favorable conditions for sulfate reducing bacteria (SRB). Ni retention was likely mainly governed by sorption mechanisms on organic matter or on Fe-oxy-hydroxides. Low temperatures, combined with salinity, may explain the lack of SRB growth in the reactors and the absence of SO42− treatment. The treatment of Cu was not affected by the low temperatures and the primary mechanism of Fe removal in the PBR-AMD reactor was precipitation in the form of oxy-hydroxides.