Immune thrombocytopenic purpura's diagnosis (ITP) is based on low platelet count and exclusion of clinical conditions rather than a specific diagnostic test. We used the reticulated platelet (RP) assay to study ITP and thrombocytopenia associated with HIV infection (HIV-ITP). Data from 96 ITP and 23 HIV-ITP patients showed low platelet counts (PC) with both high or low %RP suggesting that individuals have different degrees of thrombopoiesis. About 20% of ITP and 46% of HIV-ITP patients had %RP in the 'low' or 'normal' ranges. Grouped by platelet count <30x10(9)/L, 24% ITP and 36% HIV-ITP patients had 'low' to 'normal' %RP. The patient population did not show correlation between PC and %RP, but individuals showed an inverse relationship. Within a week of receiving IVIG, 18 ITP and 9 HIV-ITP patients' PC increased, %RP decreased. Patients with %RP measured within 24 h of IVIG treatment had lower %RP than expected, suggesting dilution by an older platelet population. ITP and HIV-ITP patients' responses to i.v. gammaglobulins were similar. Thrombopoietin levels of ITP patients did not correlate with PC, %RP, or RP count. Estimation of thrombopoiesis by RP assay provides useful information for differentiation among thrombocytopenias.