Over the last four decades, surface mining of oil sands in the boreal forest of western Canada has created large areas of disturbed land. The current regulatory framework requires that derelict land be reclaimed to pre-disturbance conditions. This has prompted the need to assess the effectiveness of reclamation, which relies on the use of salvaged materials (e.g., tailings sand and overburden), on key ecosystem components such as soil microorganisms. In this thesis, I examined landscape-scale changes in soil microbial community composition and function in response to different reclamation amendments and in natural sites comprising a regional environmental gradient. Using molecular fingerprinting (phospholipid fatty acids and denaturing gradient gel electrophoresis) and phylogenetic analyses of 16S rRNA genes, I found that microbial communities in natural soils differed from those of reclaimed soils. This dissimilarity was driven by increasing abundances of fungal and actinomycetal biomarkers in natural soils. After 30 years, however, reclamation did not place soil microbial communities on a predictable recovery path. The composition of microorganisms was particularly affected by tailings sand-based amendments. Functional potential, determined with assays targeting the activities of enzymes responsible for macromolecule degradation, was mainly impacted by prescriptions containing overburden. Variance partitioning analyses indicated that microbial responses to reclamation were partially determined by vegetation cover development, soil pH, and the fungal-to-bacterial biomass ratio. pH effects on bacterial composition were partly driven by the abundance of Acidobacteria. The relative abundances of several bacterial biomarkers covaried with individual enzyme activities, suggesting certain sub-sets of the microbial communities were functionally relevant. I tested this idea experimentally by assembling a laboratory-scale reciprocal transplant of microorganisms sourced from