Nicotine is an important constituent of tobacco and tobacco smoke. Because of this, numerous analyses of this alkaloid have been developed. Procedures for measurement of leaf and smoke nicotine are well established, and their accuracy and reliability are well known. Analyses of nicotine deposited on cellulose acetate filters are less well developed. However, these measurements are important for studying filtration efficiency and human smoking. Basically, two general procedures are known. These include steam distillation or solvent extraction of filters. Both methods require removing nicotine from the physically intact filter. They are time consuming and sometimes unreliable. A major problem is that freshly smoked filters give higher nicotine values than those which have aged between smoking and analysis. Two reasons were proposed for this difference : Either nicotine is unstable on filter material or it migrates into the cellulose acetate fiber and becomes more difficult to extract upon aging. If migration into the filter tow is the problem, an analysis which dissolves the cellulose acetate would give more consistent results. A procedure has been developed for filter analysis whereby the cellulose acetate is dissolved in acetonitrile to release any trapped nicotine. Dissolving the filter eliminates time consuming steam distillation or solvent extraction steps and assures that the recovery of nicotine is complete. After the filter is dissolved, the cellulose acetate is precipitated from solution by addition of an appropriate buffer, and an aliquot of the filtered solution is analyzed by high pressure liquid chromatography (HPLC). Two methods of HPLC are described. In both cases, the separation is achieved on a cyano-bonded silica column and detection is by ultraviolet absorption at 254nm. Different mobile phases are used in the two methods. In the first procedure, a diethylamine phosphate buffer at pH 7.56 is used; in the second procedure, a dimethylamine phosphate buffer at pH 3.00 is used. Analytical results are equivalent for both chromatographic methods, but the second procedure may offer extended analytical column life. Results of a study relating the structure of the amine in the mobile phase to nicotine retention are presented. The amount of nicotine trapped on cellulose acetate filters during smoking was determined with increasing intervals between smoking and analysis. The results demonstrate that nicotine is stable on filters and previous problems of analysis were caused by difficulty in its removal from aged filters. Results are also presented from a study of the efficiency of cellulose acetate filters for the removal of nicotine from smoke at different puff volumes. Cigarettes were smoked at 20-, 35-, and 65-ml puff volumes and the amounts of filter and smoke nicotine were determined. Changes in filter efficiency are much greater between 20- and 35-ml than between 35- and 65-ml puff volumes.