Background: Malaria is a worldwide public health problem; parasites from the genus Plasmodium spp. are the aetiological agent of this disease. The parasite is mainly diagnosed by microscope-based techniques. However, these have limited sensitivity. Many asymptomatic infections are sub-microscopic and can only be detected by molecular methods. This study was aimed at comparing nested PCR results to those obtained by microscope for diagnosing malaria and to present epidemiological data regarding malaria in Colombia's Amazon department. Methods: A total of 1392 blood samples (taken by venepuncture) from symptomatic patients in Colombia's Amazon department were analysed in parallel by thick blood smear (TBS) test and nested PCR for determining Plasmodium spp. infection and identifying infecting species, such as Plasmodium vivax, Plasmodium malariae and/or Plasmodium falciparum. Descriptive statistics were used for comparing the results from both tests regarding detection of the disease, typing infecting species and their prevalence in the study region. Bearing the microscope assay in mind as gold standard, PCR diagnosis performance was evaluated by statistical indicators. Conclusion: The present study revealed great differences between both diagnostic tests, as well as suggesting high P. malariae prevalence from a molecular perspective. This differed profoundly from previous studies in this region of Colombia, usually based on the TBS test, suggesting that diagnosis by conventional techniques could lead to underestimating the prevalence of certain Plasmodium spp. having high circulation in this area. The present results highlight the need for modifying state malaria surveillance schemes for more efficient strategies regarding the detection of this disease in endemic areas. The importance of PCR as a back-up test in cases of low parasitaemia or mixed infection is also highlighted.