The study objectives were to determine (i) if pre-ovulatory luteinizing hormone (LH) surges, undetected in urine by two immunoradiometric assays (IRMA), were detectable by an ultrasensitive immunofluorometric assay (IFMA) and (ii) the influence of creatinine adjustment on the detection and timing of the urinary LH surges. Daily urine specimens were contributed by healthy 25-36 year old volunteers during 14 ovulatory menstrual cycles for an epidemiological study conducted in 1983-1985. Specimens were selected as having been previously assayed by two IRMA without consistently detecting LH surges. These urine specimens were remeasured using an IFMA and adjusted for creatinine concentration. IFMA measurements revealed unambiguous LH surges in all cycles. Adjusting IRMA urinary LH values for creatinine concentrations revealed previously undetected LH surges in four of eight cycles. Creatinine adjustment also altered the timing of IRMA and IFMA LH surges by 1-5 days. These results demonstrate an IFMA that detects pre-ovulatory LH surges in unpreserved, frozen urine from cycles where such surges were previously undetectable. Further, creatinine adjustment can markedly affect detection and timing of the onset and peak of the urinary LH surge. While our analysis suggests that this adjustment improves the validity of the LH measure, this requires further investigation.