The sequestration of erythrocytes containing mature forms of Plasmodium falciparum in the microvasculature of vital organs may cause large discrepancies between the peripheral blood parasite count and the total body parasite burden in falciparum malaria. Despite this, parasitaemia is widely used as an indicator of prognosis and response to treatment. A simple mathematical model describing the changes in circulating and sequestered parasite numbers during acute falciparum malaria is presented. The model uses two parameters only; the standard deviation (SD) of parasite age since merogony (schizogony) as as a measure of synchronicity, and a multiplication factor each 48 h asexual life cycle. The model predicts that during the rising phase of the infection the ratio of circulating to sequestered parasites is dependent largely on the synchronicity of infection rather than multiplication rate, and that in synchronous infections parasitaemias will show considerable fluctuation when the mean stage of parasite development is in the second half of the asexual life cycle. The model fitted well to serial parasite counts from 4 patients with acute uncomplicated falciparum malaria whose infections failed to respond to ciprofloxacin. All four infections were synchronous (SD < or = 4 h), and showed large fluctuations in parasitaemia over short periods related to synchronous sequestration and subsequent reinvasion following merogony. The parasite multiplication rate was determined mainly by the efficiency of merogony or merozoite invasion rather than clearance of circulating parasitized erythrocytes. This suggests that the spleen is relatively inactive during the rising phase of the infection. Quinine treatment did not prevent sequestration but did stop subsequent multiplication.(ABSTRACT TRUNCATED AT 250 WORDS)