A three-state Markov model taking into account clinical signs of malaria infections by P. falciparum is described. The three states considered are the noninfected (state 0), the infected exhibiting no clinical signs (state 1), and the infected with clinical signs (state 2). Methods for estimating the transition rates from longitudinal data are indicated. This model was used to assess the effect on children of an intervention trial on the use of mosquito nets impregnated with insecticide. The trial was conducted in West Africa (Burkina Faso) between 1985 and 1987. The analysis showed that the intervention was most effective on transition rates between state 1 and state 2.