Abstract We have elaborated an evolutionary turbulent model of the subnebula of Saturn derived from that of Dubrulle (1993, Icarus 106, 59–76) for the solar nebula, which is valid for a geometrically thin disk. We demonstrate that if carbon and nitrogen were in the form of CO and N 2, respectively, in the early subnebula, these molecules were not subsequently converted into CH 4 and NH 3 during the evolution of the disk, contrary to the current scenario initially proposed by Prinn and Fegley (1981, Astrophys. J., 249, 308–317). However, if the early subnebula contained some CH 4 and NH 3, these gases were not subsequently converted into CO and N 2. We argue that Titan must have been formed from planetesimals migrating from the outer part of the subnebula to the present orbit of the satellite. These planetesimals were relics of those embedded in the feeding zone of Saturn prior to the completion of the planet and contained hydrates of NH 3 and clathrate hydrates of CH 4. It is shown that, for plausible abundances of CH 4 and NH 3 in the solar nebula at 10 AU, the masses of methane and nitrogen trapped in Titan were higher than the estimate of masses of these components in the primitive atmosphere of the satellite. If our scenario is valid and if our turbulent model properly describes the structure and the evolution of the actual subnebula of Saturn, the Xe/C ratio should be six times higher in Titan's atmosphere today than in the Sun, while the current scenario would probably result in a quasi solar Xe/C ratio. The mass spectrometer and gas chromatograph instrument aboard the Huygens Titan probe of the Cassini mission has the capability of measuring this ratio in 2004, thus permitting us to discriminate between the current scenario and the one proposed in this report.