In a theoretical study, benzene is compressed up to 300 GPa. The transformations found between molecular phases generally match the experimental findings in the moderate pressure regime (<20 GPa): phase I (Pbca) is found to be stable up to 4 GPa, while phase II (P4(3)2(1)2) is preferred in a narrow pressure range of 4-7 GPa. Phase III (P2(1)/c) is at lowest enthalpy at higher pressures. Above 50 GPa, phase V (P2(1) at 0 GPa; P2(1)/c at high pressure) comes into play, slightly more stable than phase In in the range of 50-80 GP, but unstable to rearrangement to a saturated, four-coordinate (at C), one-dimensional polymer. Actually, throughout the entire pressure range, crystals of graphane possess lower enthalpy than molecular benzene structures; a simple thermochemical argument is given for why this is so. In several of the benzene phases there nevertheless are substantial barriers to rearranging the molecules to a saturated polymer, especially at low temperatures. Even at room temperature these barriers should allow one to study the effect of pressure on the metastable molecular phases. Molecular phase III (P2(1)/c) is one such; it remains metastable to higher pressures up to similar to 200 GPa, at which point it too rearranges spontaneously to a saturated, tetracoordinate CH polymer. At 300 K the isomerization transition occurs at a lower pressure. Nevertheless, there may be a narrow region of pressure, between P = 180 and 200 GPa, where one could find a metallic, molecular benzene state. We explore several lower dimensional models for such a metallic benzene. We also probe the possible first steps in a localized, nucleated benzene polymerization by studying the dimerization of benzene molecules. Several new (C(6)H(6))(2) dimers are predicted.