AIMS--To examine whether a therapeutic dose of ultrasound waves, when directed through the thoracic wall to the spleen, would significantly affect the platelet count in patients with stable immune thrombocytopenic purpura (ITP). METHODS--Continuous ultrasound at 1 W/cm2 spatial average-time average (SATA) intensity for up to one minute/5 cm2 treatment field was well tolerated in 13 patients with ITP and one with non-Hodgkin's lymphoma. Five healthy controls were also similarly treated. Peak platelet increments occurred four to eight hours after ultrasound treatment in the ITP group (n = 16 treatments). RESULTS--The mean peak platelet increment was 6.25 x 10(9)/l with a 5% confidence interval of the mean (95% CI) of 3.32 to 8.93 x 10(9)/l (p = 0.0004). The mean peak platelet increment of normal controls was 6.6 (n = 5; 95% CI = -2.3 to 15.5; p = 0.21) and for sham treated patients it was 0.66 (n = 11; 95% CI = -1.5 to 2.8; p = 0.60). There was a significant inverse correlation between patient age in the ITP group and peak platelet increment (r = -0.60; p = 0.015). CONCLUSIONS--Splenic ultrasound is a novel approach to the treatment of ITP, and may find a place in its diagnosis or management.