Almost all species adapt their life-history or behaviour in the face of continuous environmental challenges. Two distinct adaptive responses can be favoured under these conditions: a phenotypic plastic response and an evolutionary response. In case of a phenotypic plastic response, selection favoured the ability of individuals to adapt by adjusting metabolic processes or specific behaviours. Adaptation is thus determined at the transcriptomic or proteomic level. In case of an evolutionary response, selection acts on the genetic composition of the population as a consequence of differential mortality and reproductive success. Adaptation is thus manifested at the DNA level. In both responses, the options for fitness maximization are constrained by trade-offs such as increasing stress resistance at the cost of offspring number. But so far it is unknown 1) which genes are the target of natural selection and how they contribute to phenotypic variation, and 2) how trade-offs are regulated in natural populations under different environmental regimes. We will address these questions by studying natural populations of C. elegans. It is a cosmopolitan nematode displaying high local genetic diversity, accessible to molecular high-throughput screens and allows for associations between genotype and phenotype to pinpoint the loci responsible for phenotypic variation. By taking advantage of the model C. elegans we are one of the first initiatives to combine phenotypic screens, association mapping and RNAi-mediated gene validation and unravel the molecular architecture of adaptation in natural populations.