Background: Several cross-sectional studies have investigated the incidence of urinary Congored dye positivity in women with preeclampsia (PE), compared to unaffected pregnancies, and reported very high sensitivity and low false positive rate in the diagnosis of PE. Objective: To determine the performance of the urinary Congo-red dot paper test at 35-37 weeks' gestation in the prediction of delivery with PE at 2 and >2 weeks after assessment. Methods: This was a prospective observational study in women attending for a routine hospital visit at 35 þ0 to 36 þ6 weeks' gestation in a maternity hospital in England. Urine samples were collected and the Congo-red dot paper test was used to assess the degree of Congo-red dye positivity. The test uses a scoring system from 1 to 8 and the higher the score the greater the degree of Congo-red dye positivity. We examined and compared the degree of Congo-red dye positivity in the groups that delivered with PE at 2 and >2 weeks with those that remained normotensive. Reproducibility was assessed by examining the inter-and intra-observer reliability of scoring on stored images with the researchers blinded to previous results. Results: The study population of 2140 women included 46 (2.1%) that subsequently developed PE (2.1%). The urinary Congo-red dot test was positive in 8.3% (1/12) and 2.9% (1/34) that delivered with PE at 2 and >2 weeks from assessment and in 0.2% (4/2094) of the unaffected pregnancies when the cutoff for Congo-red dye positivity was !5. The respective values when the cutoff used was !3 were 66.7%, 23.5%, and 16.5%, respectively. The intraclass correlation coefficient for the inter-observer reliability was 0.926 (95% CI 0.890-0.953, p<.0001) and Cohen's kappa coefficient for the intra-observer reliability was 0.904, p<.0001. Conclusions: The performance of the urinary Congo-red dot paper test at 35-37 weeks' gestation in the prediction of PE is very poor.