In patients suspected of having an ectopic pregnancy, because of a positive beta-human chorionic gonadotropin (beta-hCG) test and a sonographically undetectable gestational sac, even laparoscopy cannot always achieve a precise diagnosis, when the gestational age is very early. This means that the patient has to be followed by repeated beta-hCG measurements and repeated laparoscopy, which are expensive and inconvenient for the patient. This paper reports on a method that is simple to perform and that will provide the clinician with valuable information about the strong likelihood of a tubal pregnancy when ultrasound and beta-hCG measurements raise the suspicion but are not diagnostic. A prerequisite for this approach is the availability of a minimal volume (1-2 ml) of peritoneal fluid in the pouch of Douglas. A total of 131 patients who had a positive beta-hCG test and an undetectable gestational sac within the uterine cavity and who had a small volume of cul-de-sac fluid were studied The fluid was obtained transvaginally by means of an ultrasound-guided automatic puncture system. The beta-hCG level was determined in the cul-de-sac fluid and in the maternal serum, for comparison. In two cases with tubal pregnancy, concentrated viscous blood samples were aspirated and beta-hCG concentrations could not be measured for technical reasons, In 129 cases, the ratio of levels of beta-hCG between serum and cul-de-sac fluid could be measured In 42 out of 44 patients with intrauterine pregnancy, the ratio was greater than unity, in 82 out of 85 patients with a tubal pregnancy, the ratio was less than unity. With this ratio used for detection of tubal pregnancy, the diagnostic sensitivity of the test rr as 95.4% and the specificity was 95.2%. The positive predictive value was 97.6% and the negative predictive value 91.3%. We conclude from these findings that the measurement of the maternal serum to cul-de-sac fluid beta-hCG ratio facilitated the early diagnosis of an ectopic pregnancy in the Fallopian tube.