Currently, genetic variation is probably the most important basic resource for plant biology. In addition to the variation artificially generated by mutants in model plants, naturally occurring genetic variation is extensively found for most species, including Arabidopsis. In many cases, natural variation present among accessions is multigenic, which has historically hampered its analysis. However, the exploitation of this resource down to the molecular level has now become feasible, especially in model species like Arabidopsis, where several genes accounting for natural variation have already been identified. Dissecting this variation requires first a quantitative trait locus (QTL) analysis, which in Arabidopsis has proven very effective by using recombinant inbred lines (RILs). Second, identifying the particular gene and the nucleotide polymorphism underlying QTL is the major challenge, and is now feasible by combining high-throughput genetics and functional genomic strategies. The analysis of Arabidopsis natural genetic variation is providing unique knowledge from functional, ecological, and evolutionary perspectives. This is illustrated by reviewing current research in two different biological fields: flowering time and plant growth. The analysis of Arabidopsis natural variation for flowering time revealed the identity of several genes, some of which correspond to genes with previously unknown function. In addition, for many other traits such as those related to primary metabolism and plant growth, Arabidopsis QTL analyses are detecting loci with small effects that are not easily amenable by mutant approaches, and which might provide new insights into the networks of gene regulation.