OBJECTIVE: To assess the impact of HIV infection on the reliability of the first-trimester screening for Down syndrome, using free beta-human chorionic gonadotrophin, pregnancy-associated plasma protein-A and fetal nuchal translucency, and of the second-trimester screening for neural tube defects, using alpha-fetoprotein. PATIENTS AND METHODS: Multicentre study comparing the multiples of the median of markers for Down syndrome and neural tube defect screening among 214 HIV-infected pregnant women and 856 HIV-negative controls undergoing a first-trimester Down syndrome screening test, and 209 HIV-positive women and 836 HIV-negative controls with a risk evaluation for neural tube defect. The influence of treatment, chronic hepatitis and HIV disease characteristics were also evaluated. RESULTS: Multiples of the median medians for pregnancy-associated plasma protein-A and beta-human chorionic gonadotrophin were lower in HIV-positive women than controls (0.88 vs. 1.05 and 0.84 vs. 1.09, respectively; P < 0.005), but these differences had no impact on risk estimation; no differences were observed for the other markers. No association was found between HIV disease characteristics, antiretroviral treatment use at the time of screening or chronic hepatitis and marker levels. CONCLUSION: Screening for Down syndrome during the first trimester and for neural tube defect during the second trimester is accurate for HIV-infected women and should be offered, similar to HIV-negative women.