O-(2-Hydroxyethyl)cellulose ( 1) as formed by the alkali-catalyzed addition of ethylene oxide to cellulose flock in a slurry process is not uniformly substituted. Most of the ethylene oxide adds to HO-6 in a chain-fashion, so that ∼20% of the d-glucose residues remain totally unsubstituted at 2.0 molar substitution. Consequently, an aqueous solution thickened by 1 is highly susceptible to enzymic degradation. Stepwise decrease in concentration of alkali during etherification gives improved stability to enzymic degradation associated with a more-uniform distribution of hydroxyethyl groups between the three hydroxyl groups of the glucose residues in cellulose. The relative reactivities of hydroxyl groups and patterns of substitution were established by matching the distribution patterns from a stochastic computer-model with the distribution of substituents as determined by chemical analyses [namely hydrolysis with sulfuric acid to determine the percentage of unsubstituted glucose residues (u- 2) and with periodate oxidation for determining the percentage of unsubstituted 2,3-vicinal diol groups per residue]. The reactivities of the three hydroxyl groups at various alkali concentrations in a heterogeneous, slurry-addition process approximate those observed under homogeneous conditions, wherein the reactivities have been determined by tedious chromatographic analyses. In the variable-alkali procedure for addition of ethylene oxide, the amount of water available to the cellulose matrix in the low-alkali ( m) sequence is important both for the stability to enzymic degradation and for obtaining gel-free, thickened, aqueous solutions. Optimal stabilities and gel-free solutions are observed at intermediate water: cellulose ratios of 1.10–1.23. At a ratio of 1.23, the stability to enzymic degradation is less sensitive to percentage variations of u- 2 than in 1 prepared at higher or lower water: cellulose ratios. Although the initial degree of degradation between 1 of high molar substitution prepared at 6.8 m alkali concentration and a similar product prepared under variable alkali conditions may be related to percentage differences of u- 2, the rate and final degree of degradation do not relate to percentage differences of u- 2. An adequate interpretation, utilizing known cellulase turnover-rates, is found in stochastic-model projections of the distribution of consecutive 2 residues not substituted at HO-2. The results indicate that ( 1) more-uniform substitution through equalization of hydroxyl reactivities is achieved by lowering the alkali concentration, and ( 2) more-uniform substitution of 2 of the many cellulose-chains being substituted is achieved by employing an optimal amount of “available” water during etherification.