This study evaluated the immunochromatographic (IC) capillary flow technology for detection of T. vaginalis antigens in vaginal and urine specimens. The IC antigen-detection test and microscopy wet smears were assessed against the InPouch TV culture. Vaginal-swabs and first-voided urine specimens were obtained from 348 sexually active symptomatic or asymptomatic, > or =16 years old women. Nineteen samples (5%) were positive by culture (95%, CI 0.3-0.8). The IC test (vaginal-specimens) was more sensitive than the wet smears. IC sensitivity was 84% (95%CI, 0.68-1.01), specificity 98% (95%CI, 0.96-0.99), positive predictive value (PPV) 70% (95%CI, 0.51-0.88), negative predictive value (NPV) 99% (95%CI, 0.98-1), (positive likelihood ratio (+LR): 39.6, and negative likelihood ratio (-LR) 0.2. The IC test using urine as a substrate had less performance than both microscopy and IC test of vaginal specimens. Urine-IC sensitivity was 53% (95%CI, 0.30-0.75), specificity 99% (95% CI, 0.98-1.00), PPV 77% (95%CI, 0.54-1), NPV 97% (95%CI, 0.96-0.99), +LR: 57.7, and -LR: 0.5. Sensitivity of vaginal smear was 68% (95%CI, 0.48-0.89), specificity 100%, PPV 100% (95% CI, 1), NPV 98% (95% CI, 0.97-1), +LR>225, & -LR: 0.3. T. vaginalis was diagnosed in two wet urine samples, but not in smears. The IC antigen improved T. vaginalis diagnosis, especially in screening, rapid, or point-of-care test, but in urine was less reliable than with vaginal smear.