Abstract
Fluctations in luteinizing hormone are believed to consist of irregularly spaced sharp increases separated by periods of exponential decay. A simple method is presented for analysing such fluctuations when the data consist of uniformly sampled observations of hormone. Specific allowance for the exponential decay in the absence of pulses is made via a time series model before assessing the number and extent of pulses. All calculations are done using MINITAB regression programs. The results have been compared with those obtained by three established models and are in general agreement.