Recombinant Ralstonia eutropha capable of producing poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) copolymer [P(3HB-co-3HHx)] from fructose was engineered by introduction of genes for crotonyl-CoA reductase (CCR) from Streptomyces cinnamonensis (ccr(Sc)) and for PHA synthase and (R)-specific enoyl-CoA hydratase from Aeromonas caviae (phaC-J(Ac)). In this recombinant strain, C-6-acyl-CoA intermediates were provided via beta-ketothiolase-mediated elongation of butyryl-CoA, which was generated from crotonyl-CoA by the function of CCR. The recombinant strain could accumulate the copolyester up to 48 wt % of dry cell weight with 1.5 mol % of 3HHx fraction from fructose, when the expression of ccr(Sc) under the control of the P-BAD promoter was induced with 0.01% L-arabinose. The absence Of L-arabinose or the deletion of ccr(Sc) from the plasmid resulted in accumulation of poly(3-hydroxybutyrate) homopolymer, indicating the critical role of CCR in the formation of the 3-hydroxyhexanoate unit. Higher CCR activity obtained by the addition of a larger amount Of L-arabinose did not affect the composition but reduced the intracellular content of the copolyester. The P(3HB-co-1.5 mol % 3HHx) copolyester produced from fructose by the recombinant R. eutropha showed relatively lower melting temperatures (150 degreesC and 161 degreesC) and lower crystallinity (48 +/- 5%) compared to those (175 degreesC and 60 +/- 5%) of P(3HB) homopolymer. It has been found that the incorporation of a small amount (1.5 mol %) of 3HHx units into P(3HB) sequences leads to a remarkable change in the solid-state proper-ties of P(3HB) crystals. The present study demonstrates the potential of the engineered pathway for the production of copolyesters having favorable characteristics from inexpensive carbon resources.