In order to get the optimum conditions for the counting of C^(14) activity in liver lipid samples, the count rate, counting efficiency and the accuracy of counting were investigated through the gains of 300-1000, using internal standard method and external standard channels ratio method. The activities of samples were counted through the two channels (windows) : C^(14) wide range (channel 1) and C^(14) over H³ narrow-range (channel 2). In internal standard method, channel 1 and gain 500 were selected as optimum instrumental conditions, and 30 ㎎ lipid sample showed more than 92% counting efficiency and less than +0.9 % deviation as a optimum sample condition. When one use this method of standardisation, the 30 ㎎ sample amount should be kept in his experimental counting, because there would be much higher deviation from theoretical activity in other amounts of sample, either more or less. In external standard channels ratio, the difference of accuracies on both channels were not so much, however, the counting in channel 2 was sightly more accurate. When petroleum ether-involving samples were used, 50 ㎎ lipid level showed best results, and gain 850 was selected as a optimum condition. Those conditions would represent more than 59.9% counting efficiency and less than +1.0% deviation. If this method is used, it will be ideal to make the standard curve of efficiency vs. channels ratio using the Name range of sample and same condition of instrument as those of real experiment.