The existence of the RNA world, in which RNA acted as a catalyst as well as an informational macromolecule, assumes a large prebiotic source of ribose or the existence of pre-RNA molecules with backbones different from ribose-phosphate. The generally accepted prebiotic synthesis of ribose, the formose reaction, yields numerous sugars without any selectivity. Even if there were a selective synthesis of ribose, there is still the problem of stability. Sugars are known to be unstable in strong acid or base, but there are few data for neutral solutions. Therefore, we have measured the rate of decomposition of ribose between pH 4 and pH 8 from 40°C to 120°C. The ribose half-lives are very short (73 min at pH 7.0 and 100°C and 44 years at pH 7.0 and 0°C). The other aldopentoses and aldohexoses have half-lives within an order of magnitude of these values, as do 2-deoxyribose, ribose 5-phosphate, and ribose 2,4-bisphosphate. These results suggest that the backbone of the first genetic material could not have contained ribose or other sugars because of their instability.The discovery of catalytic RNA (1, 2) gave credibility to previous suggestions that the first living organisms were RNA molecules with catalytic activity (3-7), a concept known as the RNA world (8, 9). We take the RNA world to mean the period before the evolution of protein synthesis when RNA itself contained the genetic information and acted as a catalyst in biochemical reactions. This RNA would have contained a ribose-phosphate backbone with adenine, uracil, guanine, and cytosine (A, U, G, and C) as the bases. It is likely that there was a period before the RNA world, which we will call the pre-RNA world, when living organisms contained a backbone different from ribose-phosphate and possibly bases different from A, U, G, and C. These molecules would still have encoded genetic information and acted as catalysts.There are a number of reasons to believe that the original backbone was not ribose-phosphate, although A, U, G, and C are likely candidates for the bases. First, the polymerization of formaldehyde to sugars, known as the formose or Butlerow reaction, produces almost all possible pentoses and hexoses, including branched-chain sugars, with no selectivity for ribose (10, 11). Also, while the formose reaction works under prebiotic conditions, it is not particularly efficient (12, 13). Another problem is that nucleoside synthesis from the heating of purines with ribose is inefficient and does not work at all with pyrimidines (14, 15). Even if nucleosides could be efficiently synthesized, the products would be racemic, leading to enantiomeric cross-inhibition difficulties in template polymerizations (16).There is an additional problem as severe as any of the above-the stability of ribose. Sugars are well known to be unstable in acid and base (17-21), but there are few data for neutral solutions (22, 23). We find that ribose and other sugars have surprisingly short half-lives for decomposition at neutral pH, making it very unlikely that sugars we...