Abstract: This study evaluates the feasibility of landfarming biotreatment of petroleum-contaminated soils obtained from a sub-Arctic site at Resolution Island, Nunavut, Canada, and evaluates the changes in composition of the semi- and non-volatile petroleum hydrocarbon fractions during the biotreatment. Pilot-scale landfarming experiments were conducted in a laboratory in soil tanks under temperature profiles representative of the 3-year site air temperatures in July and August where temperature varied uniformly between 1°C and 10°C over 10d. The site soils were acidic and N-deficient, but contained indigenous populations of hydrocarbon-degrading microorganisms. Biostimulation with nitrogen and phosphorus nutrient amendments to achieve CTPH:N:P molar ratio of 100:9:1, and CaCO3 amendment at 2000mgKg−1 for maintaining neutral pH, and periodic 10-day tilling, reduced total petroleum hydrocarbon (TPH) concentrations by up to 64% over a 60-day period. The rate and extent of semi-volatile (F2: >C10–C16) and non-volatile (F3: >C16–C34) petroleum hydrocarbon fractions in the landfarms containing higher initial TPH levels (∼2000mgKg−1) and lower TPH levels (∼1000mgKg−1) were compared. Significant biodegradation of the F2 and F3 fractions occurred in both of those systems. First-order biodegradation rate constants of up to 0.019±0.001d−1 were determined for the F3 hydrocarbon fraction and were similar to the F2 fraction biodegradation rate constants of up to 0.024±0.005d−1. Biodegradation profiles of the C14, C16 and C18 alkanes revealed that at TPH concentrations above 1000mgKg−1 these compounds are degraded concurrently, whereas below 1000mgKg−1 the higher-molecular weight alkanes are preferentially degraded. After the 60-day treatment period, the TPH concentration was approximately 500mgKg−1, and the residual TPH mass was largely associated with particles and aggregated particles with diameters of 0.6–2mm, rather than the larger or finer particles and aggregates. [Copyright &y& Elsevier]