Abstract Background Humans living in regions with high falciparum malaria transmission intensity harbour multi-strain infections comprised of several genetically distinct malaria haplotypes. The number of distinct malaria parasite haplotypes identified from an infected human host at a given time is referred to as the complexity of infection (COI). In this study, an amplicon-based deep sequencing method targeting the Plasmodium falciparum apical membrane antigen 1 (pfama1) was utilized to (1) investigate the relationship between P. falciparum prevalence and COI, (2) to explore the population genetic structure of P. falciparum parasites from malaria asymptomatic individuals participating in the 2007 Demographic and Health Survey (DHS) in the Democratic Republic of Congo (DRC), and (3) to explore selection pressures on geospatially divergent parasite populations by comparing AMA1 amino acid frequencies in the DRC and Mali. Results A total of 900 P. falciparum infections across 11 DRC provinces were examined. Deep sequencing of both individuals, for COI analysis, and pools of individuals, to examine population structure, identified 77 unique pfama1 haplotypes. The majority of individual infections (64.5%) contained polyclonal (COI > 1) malaria infections based on the presence of genetically distinct pfama1 haplotypes. A minimal correlation between COI and malaria prevalence as determined by sensitive real-time PCR was identified. Population genetic analyses revealed extensive haplotype diversity, the vast majority of which was shared across the sites. AMA1 amino acid frequencies were similar between parasite populations in the DRC and Mali. Conclusions Amplicon-based deep sequencing is a useful tool for the detection of multi-strain infections that can aid in the understanding of antigen heterogeneity of potential malaria vaccine candidates, population genetics of malaria parasites, and factors that influence complex, polyclonal malaria infections. While AMA1 and other diverse markers under balancing selection may perform well for understanding COI, they may offer little geographic or temporal discrimination between parasite populations.