SummaryWe investigated the relationship between selected host haematological and parasitological parameters and the density and infectivity of Plasmodium falciparum gametocytes. 143 individuals (age range 1-62 years) attending an outpatient clinic in Farafenni, The Gambia, who had peripheral blood gametocytaemia were recruited (mean gametocyte density 123.7/l, range 5-17,000/l). Of the parameters measured, packed cell volume (PCV), reticulocyte count (RetC) and lymphocyte count (LyC) were significantly associated with gametocyte density (r ϭ Ϫ 0.17, P Ͻ 0.05; r ϭ 0.21, P Ͻ 0.01; r ϭ 0.18, P Ͻ 0.05, respectively). Data from membrane feeding experiments in which 15 or more mosquitoes were dissected showed that 60.7% (53/87) of gametocyte carriers infected one or more mosquitoes. Gametocyte density was strongly correlated with transmission success (TS) (r ϭ 0.3, P Ͻ 0.005) and, in successful infections, with both mosquito prevalence (MP) (r ϭ 0.36, P Ͻ 0.005) and mean oocyst burden (MOB) (r ϭ 0.65, P Ͻ 0.0001). None of the other factors measured were significantly associated with any of these indices in bivariate analysis. Regression modelling showed that both gametocyte density and PCV were positively associated with gametocyte carrier infectivity to mosquitoes (LR 2 ϭ 100.7 and 47.2, respectively) and, in successful infections, with MOB ( ϭ 0.16, t ϭ 4.9, P Ͻ 0.001;  ϭ 0.02, t ϭ 2.3, P Ͻ 0.05, respectively). The positive association with PCV suggests that blood meal quality influences infection probably as a nutritional requirement, however, as this effect was most apparent at high gametocyte densities, its epidemiological significance is questionable. Though the haematological parameters associated with gametocyte density are a direct consequence of asexual infection, they may also represent an adaptive mechanism for optimization of sexual stage development.