Mechanical relaxation processes in polymer melts and networks are discussed. This is performed by decomposing master curves of the dynamic shear compliance into i) glass relaxation with its plateau compliance JeN; ii) shearband process with its relaxation strength AJB, which is reciprocal to the total crosslink density pc; and iii) flow relaxation AJF and viscous flow (for uncrosslinked melts only). Plateau compliance JeN is exponentially reduced only by effective crosslinks (pc ~-pd30). This behavior is understood on the level of a meander superstructure, which includes shearbands.The observed saturation inleN at higher dicumylperoxide (DCUP) crosslinking-which doesn't appear with radiation -can be explained by the lack of chemically induced effective crosslinks across the interfaces among meander cubes. This lack may be a consequence of DCUP molecules concentrating at the interfaces and thereby preventing the contact and radical recombination between chains at adjacent meander faces.Crosslink densities Pc (per monomer), determined from the reduction of shearb and relaxation strength, vary linearly with the crosslinking agent and read: pc ~ 2,4-10 -z Dose/MGy and Pc ~ 0.97 9 10 -z DCUP/phr for radiation and DCUP crosslinking, respectively. This implies, e. g., that a dose of 0.4 MGy (40 Mrad) is equivalent to 1 part DCUP phr in a crosslinking polyisoprene. From activation-curve analysis it follows that 3 rid stays constant, and esr (flee energy of formation of a segment-dislocation) and Qr -Qyo (activation energy for segmental jumps) vary with the square of pc, as does the glass temperatur Tg -Tgo from DSC measurements.