Thrombocytopenia has been reported in the majority of malaria studies. Some but not all studies suggest the possible role of platelets in the pathology of severe malaria. We assess the association of admission platelet count with malaria complications and mortality in vivax and falciparum malaria. This is a prospective, observational study of patients aged 18 years and above admitted in a tertiary care teaching hospital from August 2004 to July 2006 in Manipal, India. Malaria was diagnosed based on clinical features along with positive Quantitative Buffy Coat method (QBC MP) or thin blood smear examination (Giemsa stain). Platelet counts were measured using Coulter LH 756 Analyser. Thrombocytopenia was defined as a platelet count <150×10(9)/l. A total of 131 consecutive patients were included. Sixty patients (46%) were infected with Plasmodium vivax and the rest with Plasmodium falciparum. Forty-six (35%) patients had non-severe and 24 (18%) had severe falciparum infection. The prevalence of thrombocytopenia was similar in vivax and falciparum malaria. Patients with severe falciparum malaria had a statistically significant lower platelet count (P = 0·01) compared to non-severe falciparum malaria. Severe malaria patients with renal failure (P = 0·02) or hyperparasitaemia (P = 0·03) had a statistically significant lower mean platelet count compared to non-severe falciparum malaria. Patients with involvement of more than one organ system had a lower mean platelet count compared to those with single organ involvement. The incidence of thrombocytopenia was similar in vivax and falciparum malaria. The admission platelet count is significantly lower in patients who have hyperparasitaemia and acute renal failure compared to patients without complications.