Numerous petroleum-geochemical analyses of deeply buried, high-rank, fine-grained rocks from ultra-deep wellbores by different investigators demonstrate that C15+ hydrocarbons (HCs) persist in moderate to high concentrations at vitrinite reflectance (R0) values of 2.0-5.0% and persist in measurable concentrations up to R0 = 7.0-8.0%, at which point the thermal deadline for C15+ HC's is finally approached. Qualitative analyses have been carried out on 1. (1) high-rank gas condensates which have been exposed to the HC-thermal-destructive phase, 2. (2) bitumens from high-temperature aqueous-pyrolysis experiments in the HC-thermal-destructive phase, and 3. (3) bitumens from high-rank, fine-grained rocks near the HC-thermal-destructive phase. These analyses clearly demonstrate that well-defined compositional suites are established in the saturated, aromatic, and sulfur-bearing aromatic HCs in and near the HC-thermal-destructive phase. On the other hand, accepted petroleum-geochemical paradigms place rigid limits on HC thermal stability: C15+ HCs begin thermal cracking at R0 values of 0.9% and are completely thermally destroyed by R0 = 1.35%; C2-C4 HC gases are thermally destroyed by R0 = 2.0% and methane is thermally destroyed by R0 = 4.0%. Furthermore, published data and observations in many HC basins worldwide support these models; for example, 1. (1) sharp basinal zonations of gas and oil deposits vs. maturation rank in HC basins and 2. (2) decreasing C15+ HC concentrations in some fine-grained rocks at ranks of R0 ??? 0.9%. The fact that observed data (C15+ HCs thermally stable to R0 = 7.0-8.0%) is so far removed from predicted behavior (C15+) HCs expected to be thermally destroyed by R0 = 1.35%) may be due to 1. (1) a lack of recognition of some important possible controlling parameters of organic matter (OM) metamorphism and too much importance given to other assumed controlling parameters; and 2. (2) assigning HC distribution patterns in petroleum basins to HC thermal cracking when such patterns may be due to other causes. In the first case, laboratory experiments strongly suggest that the presence of water, increasing fluid pressures, and closed systems (product retention) all suppress OM metamorphic reactions. Conversely, the absence of water, low fluid pressures, and open systems (product escape) all promote OM metamorphic reactions. These experiments also demonstrate that OM metamorphic reactions proceed by reaction kinetics greater than first order. Thus, the effect of geologic time appears to have been over-estimated in OM metamorphism. In the second case, the strong decreases in C15+ HC concentrations in fine-grained rocks with Type III OM over R0 = 0.9-1.35% are most probably due to intense primary migration and loss of HCs to drilling muds during the trip uphole in drilling operations. Data from coals demonstrate that these decreases in HC concentrations cannot be due to C15+ HC thermal destruction. Oil deposits are generally found at shallow depths in basins, and "dry gas" (methane ??? 98% of all HC gases) deposits are found at the greatest depths. This HC distribution pattern would be caused by methane, generated during the late stages of C15+ HC generation, flushing oil (including C2-C4 HC gases condensed into the liquid phase) out of deep basinal traps by Gussow's (1954) principle of differential entrapment. Hence, only "dry gas" deposits are left in the basin deeps. Oil emplacement processes in traps during expulsion and secondary migration could also contribute to the HC distribution pattern observed in petroleum basins. ?? 1993.