The ultrasonic degradation of poly(ethylene oxide) (PEO) of different initial molecular weights was studied at a fixed temperature. The time variation of the number average molecular weight was determined using Gel Permeation Chromatography (GPC). The ultrasonic degradation rate, κ, was assumed to be of the form, κ = kd(x - xlim)λ, where kd represents the rate coefficient, while x and x lim represent the molecular weight and limiting molecular weight, respectively. A continuous distribution model assuming midpoint chain scission and two different expressions for the rate coefficient was developed to satisfactorily model the experimental data. In the first method, the model was solved numerically and the values of kd and λ were determined by fitting the model to the experimental data. However, when λ is assumed to be unity, an analytical solution to the model was determined and fitted the experimental data for the degradation of PEO. To confirm this model, the effect of initial molecular weight on the degradation of other polymers (polyacrylamide (PAM), poly(butyl acrylate) (PBA), poly(methyl acrylate) (PMA)) was also investigated. The assumption of λ = 1 fitted the experimental data for the ultrasonic degradation of these polymers indicating that this is a reasonable approximation for the ultrasonic degradation of polymers. The effect of various solvents on the degradation of poly(ethylene oxide) was also investigated. No degradation was observed when PEO was degraded in acetone, acetonitrile or methanol but degradation occurred in mixtures of water and the above solvents. Thus, it was concluded that the degradation of PEO was dependent more on the vapor pressure of the solvent than on polymer-solvent interactions. © 2005 Elsevier Ltd. All rights reserved.