A method for determining whether a pregnant woman with an extremely low serum oestriol (ELSE) measurement of mid-trimester is carrying a fetus with steroid sulphatase deficiency or another more serious disorder is described. We undertook GC/MS analysis of steroids in random maternal urine samples and quantified oestriol, oestriol precursors (dehydroepiandrosterone (DHEA), 5-androstene-3 beta, 17 beta-diol, 16 alpha-hydroxy-dehydroepiandrosterone and 5-androstene-3 beta, 16 alpha, 17 beta-triol), pregnanediol, and five other steroids largely unaffected by pregnancy (androsterone, etiocholanolone, tetrahydrocortisol, 5 alpha-tetrahydrocortisol and tetrahydrocortisone). Thirty-two samples collected from seven normal pregnant women between the 7th and 27th week of pregnancy and 22 from individuals with ELSE were analysed. Diagnostic ratios of excreted products were developed. These included ratios of oestriol and oestriol precursors to the cumulative value for the five non-pregnancy-related steroids and ratios of oestriol and oestriol precursors to pregnanediol and to each other. Our data demonstrated high 3 beta-hydroxy-5-ene steroid excretion in all ELSE patients together with low urinary oestriol excretion, a situation only consistent with deficiency of steroid sulphatase. The normal individuals had high oestriol and low excretion of oestriol precursors. No patient in our series showed the low oestriol levels and low oestriol precursor values that would indicate a fetal adrenal abnormality as the underlying defect.