Notes: Abstract Potato (Solanum tuberosum L.) plants were transformed with “antisense” constructs to the genes encoding the α-and β-subunits of pyrophosphate: fructose-6-phosphate phosphotransferase (PEP), their expression being driven by the constitutive CaMV 35S promotor. (i) In several independent transformant lines, PFP expression was decreased by 70–90% in growing tubers and by 88–99% in stored tubers. (ii) The plants did not show any visual phenotype, reduction of growth or decrease in total tuber yield. However, the tubers contained 20–40% less starch than the wild type. Sucrose levels were slightly increased in growing tubers, but not at other stages. The rates of accumulation of sucrose and free hexoses when tubers were stored at 4° C and the final amount accumulated were the same in antisense and wild-type tubers. (iii) Metabolites were investigated at four different stages in tuber life history; growing (sink) tubers, mature tubers, cold-sweetening tubers and sprouting (source) tubers. At all stages, compared to the wild type, antisense tubers contained slightly more hexose-phosphates, two- to threefold less glycerate-3-phosphate and phosphoenolpyruvate and up to four-to fivefold more fructose-2,6-bisphosphate. (iv) There was no accumulation or depletion of inorganic pyrophosphate (PPi), or of UDP-glucose relative to the hexose-phosphates. (v) The pyruvate content was unaltered or only marginally decreased, and the ATP/ADP ratio did not change. (vi) Labelling experiments on intact tubers did not reveal any significant decrease in the unidirectional rate of metabolism of [U-14C]sucrose to starch, organic acids or amino acids. Stored tubers with an extreme (90%) reduction of PFP showed a 25% decrease in the metabolism of [U14-C] sucrose. (vii) Metabolism (cycling) of [U-14C]glucose to surcrose increased 15-fold in discs from growing antisense tubers, compared with growing wild-type tubers. Resynthesis of sucrose was increased by 10–20% when discs from antisense and wild-type tubers stored at 4° C (cold sweetening) were compared. The conversion of [U-14C]glucose to starch was decreased by about 30% and 50%, respectively. (viii) The randomisation of [1-13C]glucose in the glucosyl and fructosyl moieties of sucrose was decreased from 13.8 and 15.7% in the wild type to 3.6 and 3.9% in an antisense transformant. Simultaneously, randomisation in glucosyl residues isolated from starch was reduced from 14.4 to 4.1%. (ix) These results provide evidence that PFP catalyses a readily reversible reaction in tubers, which is responsible for the recycling of label from triose-phosphates to hexose-phosphates, but with the net reaction in the glycolytic direction. The results do not support the notion that PFP is involved in regulating the cytosolic PPi concentration. They also demonstrate that PFP does not control the rate of glycolysis, and that tubers contain exessive capacity to phosphorylate fructose-6-phosphate. The decreased concentration of phosphoenolpyruvate and glycerate-3-phosphate compensates for the decrease of PFP protein by stimulating ATP-dependent phosphofructokinase, and by stimulating fructose-6-phosphate,2-kinase to increase the fructose-2,6-bisphosphate concentration and activate the residual PFP. The decreased starch accumulation is explained as an indirect effect, caused by the increased rate of resynthesis (cycling) of sucrose in the antisense tubers.