Full-sib families of Pinus sylvestris (L.) from genetically identical parent clones, grown at tree geographic locations (64, 595 and 56-degrees-N). were tested for autumn frost hardiness. At one of the locations (64-degrees-N), maternal clones were grown in plots subjected to different site treatments. Progenies raised in a heated greenhouse with additional artificial light were induced to undergo two shoot elongation periods, separated by a period with long nights and non-hardening temperatures, followed by a hardening period. Frost hardiness was tested by exposing seedlings to -17, -19, or -21-degrees-C during the hardening period. Progenies produced at the northernmost locality (64-degrees-N) were the most hardy. The hardiness ranking among localities and the magnitude of differences in hardiness were in accordance with earlier results obtained from the same crosses that had undergone only one shoot elongation period. Effects of maternal site treatments were significant but small compared with family differences. The growth regimes used in the present study did not eliminate maternal effects induced by geographic location, but maternal effects related to site treatment decreased substantially in relation to family variation when compared with seedlings that had undergone only a single shoot elongation period. The results are discussed in relation to other studies of the same crosses raised under different growth conditions.