Thrombocytopenia in the first days of life, in association with evidence of reduced megakaryocytopoiesis and platelet production at birth, is common in sick preterm babies. Thrombopoietin (Tpo) is the major regulator of platelet production in adults. However, these babies have low Tpo levels at birth, suggesting that the Tpo response to thrombocytopenia may be impaired. To test this hypothesis we 1) measured Tpo levels, 2) measured circulating megakaryocyte progenitors serially over the first 12 d of life in 13 preterm babies with early onset thrombocytopenia and in 14 control babies with evidence of normal megakaryocytopoiesis, and 3) measured Tpo levels in thrombocytopenic children (n = 13). In control babies, platelet counts and progenitor numbers remained normal and Tpo levels were consistently low-d 1:160+/-23 pg/mL (mean+/-SEM), d 4/5: 154+/-18 pg/mL and d 12: 150+/-58 pg/mL. In thrombocytopenic babies, platelet counts and megakaryocyte progenitor numbers were significantly lower than controls at d 1: platelets 130+/-14 x 10(9)/L versus 255+/-20 x 10(9)/L (p < 0.001) and megakaryocyte progenitors 552 versus 3907 colonies/mL (mean, p < 0.001), and fell further to nadir on d 4/5: platelets 76+/-6 X 10(9)/L versus 259+/-21 x 10(9)/L (p < 0.001) and MK progenitors 479 versus 2742 colonies/mL (p < 0.05). Tpo levels were only slightly raised on d 1:247+/-52 pg/mL (p = 0.24), but then rose sharply by d 4/5: 425+/-75 pg/mL (p < 0.001). By d 12, platelet count, megakaryocyte progenitors and Tpo level (145+/-29 pg/mL) had returned to control levels. Tpo levels at platelet nadir in thrombocytopenic babies were significantly lower than in thrombocytopenic children: mean 425 versus 1383 pg/mL (p < 0.001). These data show that Tpo is important in platelet homeostasis in preterm babies, with a close reciprocal relationship with platelet count and progenitor numbers during thrombocytopenia. However, the increase in Tpo levels seen in these babies was modest, despite significantly impaired megakaryocytopoiesis, and when compared with that seen in children with thrombocytopenia. This offers further evidence that preterm babies have an impaired Tpo response to thrombocytopenia and suggests that recombinant human Tpo may have a role in the prevention/treatment of preterm thrombocytopenia.