The study was aimed at developing cellulose acetate asymmetric membrane capsules (AMCs) of acyclovir for its controlled delivery at the absorption site. The AMCs were prepared by phase inversion technique using wet process. A 2(3) full factorial design assessed the effect of independent variables (level(s) of polymer, pore former, and osmogen) on the cumulative drug release from AMCs. The buoyant optimized formulation F7 (low level of cellulose acetate; high levels of both glycerol and sodium lauryl sulphate) displayed maximum drug release of 97.88 ± 0.77% in 8 h that was independent of variation in agitational intensity and intentional defect on the cellulose acetate AMC. The in vitro data best fitted zero-order kinetics (r (2) = 0.9898). SEM micrograph of the transverse section confirmed the asymmetric nature of the cellulose acetate capsular membrane. Statistical analysis by Design Expert software indicated no interaction between the independent variables confirming the efficiency of the design in estimating the effects of variables on drug release. The optimized formulation F7 (desirability = 0.871) displayed sustenance of drug release over the drug packed in AMC in pure state proving the superiority of osmotically active formulation. Conclusively the AMCs have potential for controlled release of acyclovir at its absorption site.