The stability of free beta hCG (human chorionic gonadotrophin) was assessed using a dual alpha-fetoprotein and free beta hCG assay. A significant increase in free beta hCG concentration was observed in heparinized samples left unseparated for 24 h or more, the mean increase in 21 samples being 10.2% after 24 h (P = < 0.001), increasing to 45.7% after 96 h. Similar results were also obtained in clotted samples. The effect of the increase in free beta hCG on the Down's risk estimate was calculated to assess the impact of delayed sample transport and separation. The Down's risk increased in all samples with increasing separation time, but this was most significant in two samples where, using a cut off of one in 250, the risk classification changed from low risk to high risk. These results suggest that delayed sample separation can have a significant effect on screening programmes using free beta hCG, particularly with respect to those patients whose risk classification is changed.