A prospective comparison of the antimalarial efficacy of bed nets was conducted with 341 pregnant women living in a mesoendemic malarious area of the Thai-Burmese border. Women in 3 adjacent study sites were allocated at random to receive either a single size permethrin-impregnated bed net (PIB), a non-impregnated bed net (NIB), or to a control group who used either their own family size non-impregnated bed net (FNIB) or no net. In one study site, but not the other 2, PIB significantly reduced parasite densities and, together with FNIB, reduced the incidence of malaria in pregnancy from 56% to 33% (relative risk = 1.67, confidence interval = 1.07-2.61, P = 0.03, allowing for parity). Anaemia proved a more sensitive marker of bed net antimalarial efficacy than parasite rates. The incidence of anaemia (haematocrit < 30%) at all study sites was significantly lower at delivery in the PIB (27%) and FNIB groups (21%) than in the NIB group (41%) or those using no net (56%). This suggests that a significant proportion of the malaria in pregnancy in this mesoendemic area was sub-patent. Both patent Plasmodium falciparum parasitaemia and anaemia were associated with a reduction in birth weight. Infant mortality was high (16%) and strongly associated with prematurity, low birth weight and maternal anaemia. PIB were well tolerated and had no apparent adverse effect on the pregnancy or infant development. Although the overall effect of bed nets on patent parasitaemia was marginal, they were associated with a significant reduction in maternal malaria-associated anaemia.(ABSTRACT TRUNCATED AT 250 WORDS)