Upon Plasmodium falciparum infection of the red blood cells (RBCs), the parasite replicates and consumes haemoglobin resulting in the release of free heme which is rapidly converted to hemozoin crystallites. The bulk magnetic susceptibility of infected RBCs (iRBCs) is changed due to ferric (Fe3+) paramagnetic state in hemozoin crystallites which induce a measurable change in spin-spin relaxation (transverse relaxation) rate in proton nuclear magnetic resonance (NMR) of iRBCs. Earlier, our group reported that this transverse relaxation rate (R2) can be measured by an inexpensive, portable 0.5 Tesla bench top magnetic resonance relaxometry (MRR) system with minimum sample preparation and is able to detect very low levels of parasitemia in both blood cultures as well as animal models. However, it was challenging to diagnose malaria in human blood using MRR, mainly due to the inherent variation of R2 values of clinical blood samples, caused by many physiological and genotypic differences not related to the parasite infection. To resolve the problem of baseline R2 rates, we have developed an improved lysis protocol for removing confounding molecular and cellular background for MRR detection. With this new protocol and by processing larger volume of blood (>1 ml), we are able to reliably detect very low level of parasitemia (representing early stage of infection, ~0.0001%) with a stable baseline and improved sensitivity using the current MRR system.