Alterations of mitochondrial DNA copy number (mtDNAcn) in the blood (mitochondrial to nuclear DNA ratio) appear associated with several systemic diseases, including primary mitochondrial disorders, carcinogenesis, and hematologic diseases. Measuring mtDNAcn in DNA extracted from whole blood (WB) instead of from peripheral blood mononuclear cells or buffy coat may yield different results due to mitochondrial DNA present in platelets. The aim of this work is to quantify the contribution of platelets to mtDNAcn in whole blood [mtDNAcn(WB)] and to propose a correction formula to estimate leukocytes' mtDNAcn [mtDNAcn(L)] from mtDNAcn(WB). Blood samples from 10 healthy adults were combined with platelet-enriched plasma and saline solution to produce artificial blood preparations. Aliquots of each sample were combined with five different platelet concentrations. In 46 of these blood preparations, mtDNAcn was measured by qPCR. MtDNAcn(WB) increased 1.07 (95%CI 0.86, 1.29; p<0.001) per 1000 platelets present in the preparation. We proved that leukocyte count should also be taken into account as mtDNAcn(WB) was inversely associated with leukocyte count; it increased 1.10 (95%CI 0.95, 1.25, p<0.001) per unit increase of the ratio between platelet and leukocyte counts. If hematological measurements are available, subtracting 1.10 the platelets/leukocyte ratio from mtDNAcn(WB) may serve as an estimation for mtDNAcn(L). Both platelet and leukocyte counts in the sample are important sources of variation if comparing mtDNAcn among groups of patients when mtDNAcn is measured in DNA extracted from whole blood. Not taking the platelet/leukocyte ratio into account in whole blood measurements, may lead to overestimation and misclassification if interpreted as leukocytes' mtDNAcn.