The curve for decay of 14C in rat liver protein labelled by injection of NaH14CO3 was analysed to obtain the average turnover rate of mixed liver protein. Three different methods of analysis were used. (1) Unlike decay curves from homogeneous proteins, the curve did not fit a single exponential, but a good fit was obtained with three exponentials. By assuming that the mixture contained three major components with different turnover rates, the calculated value for the average turnover rate (k) was close to 40% per day. (2) k was also calculated from the area under the decay curve, a method which makes no assumptions about the number of proteins in the mixture. This method also gave a value close to 40% per day. (3) It was shown empirically, both by simulation of decay of label in model mixtures of protein and with the decay curve measured in vivo, that k can be calculated from the time taken for the specific radioactivity to fall to 10% of its maximum value. This is an advantage, since the other two methods require the decay curve to be measured over a much longer period of time.