Two series of experiments with Pinus sylvestris L., 10 and 20 years old, were analysed with regard to genotype-environment interaction. The series cover a wide range of site conditions with regard to length of growing season, environmental differences within localities and spacing between trees. The genetic material comes from factorial crosses in one of the series and a mixture of full-sib, half-sib and provenance material in the other. The origin of the parents is south-central Sweden, and the trials cover approximately the same geographical range as the parent trees. Despite few replicates and large experimental error, the following conclusions could be drawn: (a) genotype-environment interaction for Pinus sylvestris in south-central Sweden seems to be of importance for growth capacity, but the reaction differs between young and old plant stages; (b) the origin of the genetic material, as long as it is restricted to south-central Sweden, does not seem to have any significant effect on genotype-environment interaction; (c) a large part of the interaction seems to be explained by a linear relationship between the genotypes and an environmental index.