Sirs: We do appreciate the interest in our study, stating that there were no statistically significant differences in sensitivities between self-collected vaginal swab (SCVS) and the combined SCVS and first catch urine (FCU) as one sample. 1 Boman and Nylander highlight that different diagnostic methods may give different results. 2 They claim that our study and a Dutch study presented in a poster (but to our knowledge not yet published) 3 in 2007 at the ISSTDR/IUSTI meeting in Seattle, WA, USA indicate a difference in sensitivities according to whether strand displacement assay (SDA) or polymerase chain reaction (PCR) was used as the diagnostic method. We, however, disagree in their conclusion of the studies. In our study the sensitivities for SCVS and the combined SCVS-FCU specimens by PCR were 97.2% (105/108) and 94.4% (102/108), respectively. The corresponding outcome for SDA was 95.2% (60/63) and 96.8% (61/63), respectively. In the Dutch study when using SDA there were only two discordant samples but when using PCR there were nine. The sensitivities for SCVS and the combined SCVS-FCU specimens by PCR were 93.2% (41/44) and 88.6% (39/ 44), respectively, and by SDA were 96% (48/50) and 100% (50/50), respectively. No statistically significant differences or any clear trends were revealed in either study. This may of course have been caused by few positive tests sampled by each method, i.e. low power. Since there may be important differences in feasibility and convenience between SCVS and the combined sampling, due mainly to toilet accessibility, we do believe SCVS is a reliable sampling method and that it is, so far, the one to be recommended in a screening context, irrespective of which nucleic acid amplification test is used. We do agree that there is a need for further studies regarding optimal sampling methods when using common commercially available diagnostic methods, as well as in the necessity of investigating differences between methods for diagnosing genital Chlamydia trachomatis infections in women.