The aim of this study was to determine whether a patient's endocervical swab specimen can be transported in first void urine (FVU) as combined specimens for the detection of Mycoplasma genitalium by real-time PCR. The study also compared two different DNA extraction methods for observation of possible PCR inhibition. Three specimens, one endocervical swab specimen transported in 2-SP medium, one endocervical swab specimen transported in FVU and a FVU specimen, were collected from 329 women. All sample types underwent manual DNA extraction whereas in the DNA extraction study, 329 endocervical swab specimens transported in FVU were subjected to both manual Chelex and automated BioRobot M48 DNA extraction. A total of 100 endocervical swab specimens transported in FVU from patients PCR-negative for M. genitalium in the study were used in the PCR inhibition analysis. M. genitalium was detected in 25/329 (7.6 %) women. The endocervical swab specimens transported in 2-SP medium and transported in FVU were positive for M. genitalium in 17/25 (68 %) and 24/25 (96 %) women, respectively. The FVU specimens alone were positive for M. genitalium in 22/25 (88 %) women. In the DNA extraction study, M. genitalium DNA was detected in 24/329 (7.3 %) and 28/329 (8.5 %) of endocervical swab specimens transported in FVU subjected to manual Chelex extraction and automated BioRobot M48 extraction, respectively. Partial PCR inhibition was detected in 6 % of samples subjected to manual Chelex extraction whereas no inhibition was detected with the automated BioRobot M48 extraction. Thus endocervical swab specimens transported in FVU demonstrate higher sensitivity than FVU specimens only and have considerably increased sensitivity compared with endocervical swab specimens transported in 2-SP medium for detection of M. genitalium DNA. Moreover, automated BioRobot M48 extraction was shown to be superior to a crude manual Chelex extraction, leaving no PCR inhibition and giving a slightly higher DNA yield and/or better sensitivity.