Understanding of the mechanisms of radical production and decay in -irradiated hydrocarbon glasses, and in -irradiated and photolyzed solutions of alkyl halides in such glasses, has been extended through studies of yields and decay rates in 3-methylpentane (3MP) and 3MP-di4, with and without charge scavengers, using ESR as the detection technique. About half of the C6Hi3 radicals produced by irradiation of pure 3MP decay at 77°K by relatively fast intraspur radical-radical reaction with time-dependent firstorder kinetics. The remainder decay with second-order kinetics (k = 1.6 X 10-25 cm3 molecule-1 sec-1), by reaction following random diffusion (D e¿ 9.0 X 10-2°cm2 sec-1). Changes in resolution of the ESR spectra of trapped C6Hi3 radicals during decay and during and following photolysis reflect changes in spacing and geometry. The G value for radical production (radicals per 100 eV absorbed) in 3MP at 77°K is 3.0 ± 0.3, while GiCgDis) from the radiolysis of 3MP-di4 is 2.0 ± 0.4. GiCgHis) in 3MP-hi4 and G(CeDi3) in 3MP-di4 are lowered by electron scavengers, implying that part of the radical formation in the pure hydrocarbons results from electron capture by cations (unless the additives are unexpectedly effective as energy scavengers). Scavengeable electron yields are higher in CeDu glass than in CeH^, which is an unexpected matrix isotope effect. G(CgHi3) and GlCgDis) from radiolysis of 3MP-hi4 and 3MP-£¿i4 glasses are increased by the presence of 1 mol % hydrogen halide, implying that hot H atoms produced from the halides by dissociative electron capture abstract H (or D) from the matrix. The energetics require capture of electrons with greater than thermal energy. In 3MP-1% HI and 3MP-1% HBr systems, ~70% of the radicals decay by the CgHia + HX -* CV.Hn + X reaction, the estimated rate constant and HI diffusion coefficient for the HI system being 8.5 X 10-5 M-1 sec-1 and 1.6 X 10-19 cm2 sec-1, respectively. HC1 forms a stable complex with C6Hi3 radicals (presumably CgHis • HC1), which can be photochemically decomposed regenerating the original radicals. Alkyl radicals produced in 3MP-<¿44 by dissociative electron capture by alkyl chlorides decay faster than identical radicals produced from bromides and iodides, whereas the rates in 3MP-h44 are independent of the geminate halide ion, consistent with recent evidence that the decay mechanisms of CH3 are different in the two matrices. Photolysis of CH3I in 3MP-hi4 produces CgHis (by abstraction of H by hot CH3), but no trapped CH3 radicals. The photolysis of CH3I in 3MP-di4 yields [CHsJ/tCgDis] = ~0.02. CH3 produced photolytically in 3MP-d44 decays faster than CH3 produced by dissociative electron capture, implying that geminate recombination with I is faster than with I-.