Citronella oi l (CO), has been reported to possess a mosquito repellent action. However, itsrnapplication in topical preparations is limited by its rapid volatility.rnThe objectives of this study were to confirm its activity and reduce the rate of evaporation,rnthereby to enhance its repellency effect via microencapsulation (ME); using gelatin simplerncoacervation method. And, to formulate the optimized MCs into topical formulation. For this,rnsodium sulphate (20%) was used as a coacervating agent. The microcapsules (MCs) were finallyrnhardened by using formaldehyde (37%) as a crosslinking agent. The effects of three variables,rnstirring rate (X l ), oil loading (X2) and amount of crosslinking agent (X3), on encapsulationrnefficiency (EE), particle size (PS) and the release rate (RR) of CO were studied. Thus, responsernsurface methodology (RSM) was employed to optimize the EE (%) and PS with respect to thosernthree variables. An effort was also made to develop topical formulations for the MCs withrnreduced CO membrane permeation. For this, four different classes of ointment bases were used:rnwhite petrolatum, wool wax alcohol, hydrophilic ointment (USP) and PEG ointment (USP).rnOintments were prepared for both the MCs and the un-microencapsulated citronella oil (uCO).rnThe effects and the nature of an ointment base on the percutaneous permeation of CO were thenrninvestigated. In vitro membrane permeation of CO was evaluated in Franz diffusion cells usingrncellulose acetate membrane at 32 Â°C, with the receptor compartment containing a water-ethanolrnsolution (50:50). The receiver phase samples were analyzed by GC/MS with citronellal as arnreference standard.rnThe effect of the amount of crosslinker was insignifficant on EE (%). The effects of the stirringrnrate, oil loading and amount of crosslinking agent were significant on PS. The EE (%) and PSrnexhibited significant effect on RR. The response surface plot and contour plot for the polynomialrnequations were constructed to find the optimum area. One optimal condition for ME of CO wasrnfound to be a stirring rate of 720 rpm, CO to gelatin ratio of 0.57: 1 and formaldehyde (37%) ofrn8.15 ml. The MCs under the optimized conditions provided EE of 60% and PS of 213 [lm. Thernoptimized MCs were observed to have a sustained release profile (70% of the content wasrnreleased at the 10'" hr) with minimum initial burst effect. Differential scanning calorimetry