The primary products of n-butoxy and 2-pentoxy isomerization in the presence and absence of O-2 have been detected using pulsed laser photolysis-cavity ringdown spectroscopy (PLP-CRDS). Alkoxy radicals n-butoxy and 2-pentoxy were generated by photolysis of alkyl nitrite precursors (n-butyl nitrite or 2-pentyl nitrite, respectively), and the isomerization products with and without O-2 were detected by infrared cavity ringdown spectroscopy 20 mu s after the photolysis. We report the mid-IR OH stretch (nu(1)) absorption spectra for delta-HO-1-C4H8 center dot, delta-HO-1-C4H8OO center dot, delta-HO-1-C5H10 center dot, and delta-HO-1-C5H10OO center dot. The observed nu(1) bands are similar in position and shape to the related alcohols (n-butanol and 2-pentanol), although the HOROO. absorption is slightly stronger than the HOR. absorption. We determined the rate of isomerization relative to reaction with O-2 for the n-butoxy and 2-pentoxy radicals by measuring the relative nu(1) absorbance of HOROO. as a function of [O-2]. At 295 K and 670 Torr of N-2 or N-2/O-2, we found rate constant ratios of k(isom)/k(O2) = 1.7 (+/- 0.1) x 10(19) cm(-3) for n-butoxy and k(isom)/k(O2) = 3.4(+/- 0.4) x 10(19) cm(-3) for 2-pentoxy (2 sigma uncertainty). Using currently known rate constants k(O2), we estimate isomerization rates of k(isom) = 2.4 (+/- 1.2) x 10(5) s(-1) and k(isom) approximate to 3 x 10(5) s(-1) for n-butoxy and 2-pentoxy radicals, respectively, where the uncertainties are primarily due to uncertainties in k(O2). Because isomerization is predicted to be in the high pressure limit at 670 Torr, these relative rates are expected to be the same at atmospheric pressure. Our results include corrections for prompt isomerization of hot nascent alkoxy radicals as well as reaction with background NO and unimolecular alkoxy decomposition. We estimate prompt isomerization yields under our conditions of 4 +/- 2% and 5 +/- 2% for n-butoxy and 2-pentoxy formed from photolysis of the alkyl nitrites at 351 nm. Our measured relative rate values are in good agreement with and more precise than previous end-product analysis studies conducted on the n-butoxy and 2-pentoxy systems. We show that reactions typically neglected in the analysis of alkoxy relative kinetics (decomposition, recombination with NO, and prompt isomerization) may need to be included to obtain accurate values of k(isom)/k(O2).