Aerated cycloalkane solutions of phenothiazine (PTH) undergo pulse radiolysis to yield a transient species (λ_max 590 nm) building up after the pulse with k= 3·42 × 10⁶ l mol^–1 s^–1(for C₅H₁₀). A comparsion of the spectrum with spectra of ³PTH, PTH·⁺, and PTH·^–(or of its conjugate acid) prepared by unambiguois routes, and with literature spectra (where known) of these species, suggests that the transient is not any of these, but is more probably the neutral radical species PT·, formed by attack of cycloalkylperoxyl radicals upon the substrate. Further evidence for this is supplied by the presence of a primary deuterium kinetic isotope effect upon the rate of formation of the transient, and alternative preparations of the neutral radical are discussed. Pulse radiolysis of an aerated solution of N-methylphenothiazine (PTMe) in cyclohexane gives a different type of spectrum (λ_max.ca. 520 nm) which is formed very rapidly even after 25 ns pulses and which is characterised by comparison with spectra of species generated by known routes as the radical cation PTMe·⁺, possibly formed by electron transfer between PTMe and RO₂·. N-Benzyl-3,7-dioctylphenothiazine (dioctyl-PTCH₂Ph) again reacts with RO₂· on a ns time scale at mM solute concentrations; the species formed in this case has not been identified unambigiously, but may be dioctyl-PT· formed by the loss of the benzyl group.