Above-below ground interactions are crucial to the functioning of terrestrial ecosystems. However, little is known about the functional links between plants and soil eukaryotic microorganisms. Exploring the genes expressed by soil micro-eukaryotes through metatranscriptomics may contribute to filling this knowledge gap. We applied this method to study two contrasting alpine soils in which plant growth rates differ, owing to different snow-cover durations. Soil polyA+ RNAs from Late Snow Melt site (LSM) and Early Snow Melt site (ESM) was used to generate two libraries of ~ 40000 putative mRNA reads using the Roche 454 technology. Using a combination of annotation pipelines, 38% of these putative mRNA reads were successfully assigned to specific functions. Fungal proteins were most abundant, but plant and animal proteins were also frequently observed. Between 600 and 900 enzymes were listed for each soil. Collector’s curves indicated that sampling was partial, yet the representation of some proteins was sufficient to highlight differences between ESM and LSM metabolic pathways. Both ribosomal structural proteins and ubiquitin ligases were under-represented in ESM, suggesting a decrease in both translation and protein degradation rates. Several enzymes involved in glucose catabolism were also under-represented in ESM. These differences in gene expression point towards a slower use of cellular energy belowground in the habitats dominated by slow-growing plants. Our study thus opens up new perspectives for linking above- and belowground functional traits and for assessing the genes involved in the carbon cycle by using metatranscriptomics.