Fetal male DNA can be identified in maternal blood by polymerase chain reaction (PCR) amplification of Y-specific sequences. This technology has not reached a satisfactory accuracy and reproducibility in fetal gender determination because of the very low concentration of fetal cells. Our purpose was to evaluate the possibility of improving the reliability of this test by setting up a repeated amplification system. We amplified, by nested PCR of the Y-specific sequence DYS14, 137 DNA samples extracted from maternal peripheral blood (93 from male-bearing and 44 from female-bearing pregnancies ranging from the 6th to the 36th gestational week). Each maternal DNA sample was tested doubly, in two different PCR sessions, with a total of four amplifications. We obtained discordant results in the four amplifications in 82/137 (60%) samples. The best interpretation of these discordant results was obtained by applying a positivity cutoff of at least two positive amplifications for considering a DNA sample as belonging to a male-bearing pregnancy. We obtained a sensitivity of 83%, a specificity of 93%, a positive predictive value of 96% and a negative predictive value of 72% in fetal male gender diagnosis. By applying this quadruple testing system, we significantly improved PCR accuracy and predictive values compared with single and double testing of the same samples. We conclude that, for future investigations of fetal DNA retrieved from maternal blood, the application of a quadruple testing system is better than the single PCR test.