Naturally acquired protective immunity to Plasmodium falciparum malaria takes years to develop. It relies mainly on Abs, particularly IgG specific for Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1) proteins on the infected erythrocyte surface. It is only partially understood why acquisition of clinical protection takes years to develop, but it probably involves a range of immune-evasive parasite features, not least of which are PfEMP1 polymorphism and clonal variation. Parasite-induced subversion of immunological memory and expansion of “atypical” memory B cells may also contribute. In this first, to our knowledge, longitudinal study of its kind, we measured B cell subset composition, as well as PfEMP1-specific Ab levels and memory B cell frequencies, in Ghanaian women followed from early pregnancy up to 1 y after delivery. Cell phenotypes and Ag-specific B cell function were assessed three times during and after pregnancy. Levels of IgG specific for pregnancy-restricted, VAR2CSA-type PfEMP1 increased markedly during pregnancy and declined after delivery, whereas IgG levels specific for two PfEMP1 proteins not restricted to pregnancy did not. Changes in VAR2CSA-specific memory B cell frequencies showed typical primary memory induction among primigravidae and recall expansion among multigravidae, followed by contraction postpartum in all. No systematic changes in the frequencies of memory B cells specific for the two other PfEMP1 proteins were identified. The B cell subset analysis confirmed earlier reports of high atypical memory B cell frequencies among residents of P. falciparum–endemic areas, and indicated an additional effect of pregnancy. Our study provides new knowledge regarding immunity to P. falciparum malaria and underpins efforts to develop PfEMP1-based vaccines against this disease.