Antenatal malaria screening with a rapid diagnostic test (RDT) and treatment only of women with positive RDT findings may potentially prevent low birth weight resulting from malaria. The consequences of subpatent antenatal infections below the detection limit of RDTs are incompletely understood. In Malawi, pregnant women of any gravidity status were tested at each antenatal visit for Plasmodium falciparum, using an RDT and polymerase chain reaction analysis, and were followed until delivery. Associations between antenatal infections and delivery outcomes were assessed with Poisson regression or analysis of variance. Compared with women with no detected antenatal P. falciparum infection, women with positive RDT findings delivered babies with a lower mean birth weight (2960 vs 2867 g; mean difference, -93 g [95% confidence interval {CI}, -27 to -159]; P = .006); this was not observed among women with only subpatent infections (mean birth weight, 3013 g; mean difference, 54 [95% CI, -33-140]; P = .2268). These differences were apparent early in pregnancy, during the second trimester: compared with uninfected women, women with positive RDT findings delivered babies with a lower mean birth weight (mean difference, -94 g [95% CI, -31 to -156]; P = .003), but women with subpatent infections did not (mean difference, 36 g [95% CI, -49-122]; P = .409). Subpatent antenatal P. falciparum infections were not associated with adverse delivery outcomes. The association of patent infections at enrollment with low birth weight suggests the importance of preventing P. falciparum infection early in pregnancy.