Immunization against the human platelet antigen (HPA)-1 alloantigen is the most common cause of severe fetal and neonatal thrombocytopenia. Fetal therapy has substantial risks and its indications need better definition. Of 24,417 consecutive pregnant women, 618 (2.5%) were HPA-1a negative of whom 385 entered an observational study. All were HLA-DRB3*0101 genotyped and screened for anti-HPA-1a. Their partners and neonates were HPA-1 genotyped and the latter were assessed by cord blood platelet counts and cerebral ultrasound scans. Anti-HPA-1a was detected in 46 of 387 pregnancies (12.0%; 95% CI 8.7%-15.2%). All but one were HLA-DRB3*0101 positive (odds ratio 140; 95% CI 19-1035; P< .00001). One baby died in utero, and of 26 HPA-1a-positive babies born to women with persistent antenatal antibodies, 9 were severely thrombocytopenic (8 with a count <10 x 10(9)/L, 1 with a large porencephalic cyst), 10 were mildly thrombocytopenic, whereas 7 had normal platelet counts. Severe thrombocytopenia was significantly associated with a third trimester anti-HPA-1a titer >/= 1:32 (P = . 004), but was not observed in babies of women with either transient or postnatal-only antibodies. HPA-1a alloimmunization complicates 1 in 350 unselected pregnancies, resulting in severe thrombocytopenia in 1:1,200. HPA-1a and HLA-DRB3*0101 typing combined with anti-HPA-1a titration allows selection of the majority of pregnancies at risk of severe thrombocytopenia.