The highly selective catalytic epoxidation of terminal alkenes by the complex W-VI/P-V/H2O2/CHCl3/PTC (PTC = phase transfer catalyst) system (Ishii-Venturello chemistry) has been extensively investigated by groups in several countries and recently commercialized, yet little is known with certainty about the mechanism. The substrate conversions and epoxide selectivities observed under biphasic conditions, aqueous H2O2/alkene in CHCl3, with 21 polyoxometalates (cetylpyridinium chloride as the phase-transfer catalyst, PTC) including the Ishii precursor complex, [PW12O40](3-), clearly indicate that only [PW12O40](3-) and [PW11O39](7-), which both rapidly form {PO4[WO(O-2)(2)](4)}(3-), 1, are effective. Simultaneous monitoring of organic oxygenated products and gaseous products (nearly all O-2) with several of these polyoxometalates confirm that H2O2 disproportionation is by far the dominant side reaction with several d-electron transition metal-substituted polyoxometalate catalyst precursors.Analysis of the (2)J(W-P) coupling satellites in the P-31 NMR spectra of the polytungstophosphate products from the stoichiometric reaction of 1 with alkene substrates as a function of cation, solvent, field strength, and time indicates that both a PW4 and a PW3 specie are formed initially and one PW2 specie subsequently. Several lines of netic and spectroscopic evidence indicate that two processes dominate over ah others during Ishii-Venturello epoxidation: a slow epoxidation, 1 (PW4) + alkene --> PW4, PW3, and PW2 (henceforth called "subsequent peroxo species" or SPS) + epoxide, followed by a rapid regeneration of 1 with H2O2. First, little epoxidation is observed until 1 is in appreciable concentration. Second, the rate law for epoxidation of 1-octene by the Arquad salt of 1, Arq1 (Arquad = [(C18H37)(75%) + (C16H33)(25%)](2-) [CH3]N-2), in CHCl3 at 23 degrees C is v(0) = k[1][1-octene]. Third, 1 is the dominant polytungstophosphate present under steady state turnover conditions. Fourth, the ratio of the initial rates of epoxidation is v(0(ArqSPS))/V-0(Arq1) = 0.13 +/- 0.01. Fifth, the dominant inorganic product in the formation of 1, {[WO(O-2)(2)(H2O)]O-2}(2-), is two orders of magnitude slower in alkene epoxidation than Arq1 under identical conditions at both 23 and 60 degrees C. Additional P-31 NMR studies address both ion pairing effects and dynamic exchange in 1 and the SPS PW4. A Linear correlation was found between the change in both chemical shift and (2)J(W-P) coupling constant for the SPS PW4 specie but not the SPS PW2 or the SPS PW3 species as a function of reaction time.This is consistent with the SPS PW4 specie undergoing rapid dynamic exchange on the P-31 NMR time scale. Addition of 1 equiv of 1,2-epoxybutane to tetra-n-hexylammonium SPS (THASPS) does shift the SPS PW4 resonance to high field with a larger (2)J(W-P) coupling constant in accord with the correlation. Consequently, the dynamics of SPS PW4 may reflect Rapid catalyst inactivation despite being one of two success limiting features of Ishii-Venturello epoxidation was not addressed in any previous work. Under the typical biphasic reaction conditions, catalysis nearly stops after 500 turnovers. The effects of alkene, H2O, and epoxide product on epoxidation rates and polytungstophosphate speciation monitored by P-31 NMR establish that epoxide, but apparently neither alkene nor H2O, leads to irreversible catalyst inactivation.