Ovarian function in women may be monitored by a simple, solid-phase, chemiluminescence immunoassay for the measurement of oestrone-3-glucuronide in diluted urine. An IgG fraction of antiserum to oestrone-3-glucuronyl-6-bovine serum albumin is passively adsorbed to the walls of polystyrene tubes. Oestrone\x=req-\ 3-glucuronyl-6-aminoethyl-ethyl-isoluminol is the labelled antigen. Daily samples of early morning urine are diluted in buffer (1:100; v/v), and 100 \g=m\lremoved, in duplicate, for assay. After the binding reaction (1 h at 4\s=deg\C), the solution is removed by aspiration. The antibody-bound fraction is washed once with buffer (400 \ g=m\ l ). Sodium hydroxide (2n, 200 \g=m\l ) is added and the mixture incubated for 60 min at 22\s=deg\C.Luminescence is initiated by oxidation of the label with microperoxidase/hydrogen peroxide and the signal integrated for 10s. An evaluation of the method gave the following values: sensitivity of calibration curve 3.12 \m=+-\0.75 pg/tube (mean \m=+-\ sd). The intra-assay precision (CV%) was 9.52 (20 replicates; 38.4 \ m=+-\3.66 nmol/l) and 8.61 (15 replicates; 102.4 \ m=+-\ 8.82 nmol/l). The inter-assay precision (CV%) was 11.9 (mean of 4 pools\m=-\7. 03,23.16, 52.11 and 117.53 nmol/l over 2 months). The mean bias was \m=-\0.78% over the range 28 to 448 nmol/l and different aliquots (equivalent to 0.25 to 4 \g=m\l ) of daily early morning urine gave results that were in good agreement. The concentration (nmol/l; mean \ m=+-\sd) of oestrone-3-glucuronide in samples of early morning urine collected from healthy women during the early follicular, periovulatory and luteal phases of the menstrual cycle were 40.2 \m=+-\9.9, 102.3 \ m=+-\ 39.4 and 84.3 \m=+-\13.3 nmol/l, respectively. In addition, the results obtained from the analysis of EMU throughout 6 complete menstrual cycles were in good agreement with the values derived by radioimmunoassay (r = 0.94).Numerous methods have been devised to investi¬ gate ovarian function and to monitor or predict variations in potential fertility in the human female. These have been based on: i) the secretion of pituitary and ovarian hormones (WHO 1980); ii) their biochemical or biophysical effect on re¬ sponsive tissue (Moghissi 1980); or iii) the change in ovarian morphology as observed with ultrasound (Queenan et al. 1980). Consequently, a greater understanding of the interaction and temporal relationships of the many parameters associated with ovulation has been achieved.The results from studies of ovarian function and steroid metabolism have suggested that the prin¬ cipal urinary metabolites of oestradiol and proge¬ sterone might be used as indices of the start and finish of the probable fertile period ). Accordingly, methods based upon the prin¬ ciples of radioimmunoassay have been developed to measure the main oestrogen glucuronides and pregnanediol-3a-glucuronide in diluted urine (Samarajeewa & Kellie 1975; Samarajeewa et al. 1979). Subsequently, several investigators have undertaken the study of facets of this bochemical approach...