Understanding the mechanisms of drug resistance in Plasmodium vivax, the parasite that causes the most widespread form of human malaria, is complicated by the lack of a suitable long-term cell culture system for this parasite. In contrast to P. falciparum, which can be more readily manipulated in the laboratory, insights about parasite biology need to be inferred from human studies. Here we analyze the genomes of parasites within 10 human P. vivax infections from the Peruvian Amazon. Using next-generation sequencing we show that some P. vivax infections analyzed from the region are likely polyclonal. Despite their polyclonality we observe limited parasite genetic diversity by showing that three or fewer haplotypes comprise 94% of the examined genomes, suggesting the recent introduction of parasites into this geographic region. In contrast we find more than three haplotypes in putative drug-resistance genes, including the gene encoding dihydrofolate reductase-thymidylate synthase and the P. vivax multidrug resistance associated transporter, suggesting that resistance mutations have arisen independently. Additionally, several drug-resistance genes are located in genomic regions with evidence of increased copy number. Our data suggest that whole genome sequencing of malaria parasites from patients may provide more insight about the evolution of drug resistance than genetic linkage or association studies, especially in geographical regions with limited parasite genetic diversity.