We studied glycocalicin (GC), expressed as plasma GC concentration and as GC index (ratio to platelet count), in 129 thrombocytopenic patients (platelet count < 100 x 10(9)/l) and 60 sex- and age-matched controls. Seventy-two patients had idiopathic immune thrombocytopenia, 32 secondary immune thrombocytopenia, 8 microangiopathic thrombocytopenia and 17 thrombocytopenia secondary to bone marrow aplasia. Patients with immune thrombocytopenia (ITP) were also subclassified, according to their clinical behaviour, as having active disease or being in spontaneous or therapy-induced partial remission. A significant correlation was found between glycocalicin levels and platelet count both in normals and in patients with bone marrow aplasia (r = 0.75). ITP patients showed a GC index significantly higher than controls (6.02+/-7.87 vs. 0.9+/-0.2, p<0.001). When ITP patients with similar platelet count (30-50 x 10(9)/l) were studied, the mean level of GC and the GC index were significantly higher in those patients with active disease than in those in remission (0.97+/-0.38 vs. 0.58+/-0.17 microg/ml, p <0.05; 6.41+/-2.64 vs. 3.44+/-0.94, p<0.05, respectively). A longitudinal study performed in 10 patients with different subtypes of ITP suggested a positive correlation between GC index and the activity of the disease. The GC value and GC index were significantly higher in patients with microangiopathic thrombocytopenia than in controls (1.44+/-0.73 vs. 0.8+/-0.16 microg/ml, p < 0.01; and 18.77+/-22.23 vs. 0.9+/-0.2, p<0.001, respectively). The GC value was significantly lower in bone marrow failure (0.15+/-0.04 microg/ml, p<0.01) compared to controls, while no difference was observed in the GC index. Our data confirm that the GC index is helpful in differentiating thrombocytopenia due to increased platelet destruction from the one due to impaired production. In addition, the assay has been proven useful in the differential diagnosis of different ITP subtypes and their follow-up.