A randomized prophylactic drug trial was conducted in a malaria holoendemic area, in the Muheza District of Tanzania. Of 327 pregnant women, 124 received proguanil (PROG), 113 chloroquine (CQ), 90 the proguanil and chloroquine combination (CQ+PROG). Prophylaxis was supervised. Chemosuppressive efficacy was measured through the incidence of malaria breakthrough parasitaemias and clinical episodes. Groups were comparable by age, parity, residential area, and enrollment gestational age. Compliance and drug bio-availability was good. The median breakthrough time of the first parasitaemia episode for primigravidae (PG) and multigravidae (MG) was significantly shorter for the CQ group (PG = 56, MG = 78 days) than in the PROG (PG = 97, MG = 112 days) and the CQ+PROG (PG = 138, MG = 140 days) groups. 56% of the CQ group experienced 2 or more clinical episodes compared to 19% (PROG) and 10% (CQ+PROG). PROG and CQ+PROG did not differ significantly. Parasite densities and in vitro tests suggested that CQ selected for more and highly resistant strain(s). Proguanil is suitable for prophylaxis. However, proguanil resistance should be monitored as well as controlled drug distribution and usage. Malaria control strategies other than chemoprophylaxis should be investigated.