The human erythropoietin (hEPO) is the main hormone concerned with stimulation of erythropoiesis and maintain of erythroid stem cells physiologic levels. Its deficiency causes severe anemia and happens to patients with kidney failure because of traumas, graft, chemotherapy, diabetes, etc. Hormone supply is an essential treatment on these patients. Currently, the recombinant human erythropoietin (rhEPO) is mainly produced in genetically modified CHO cells, regardless of the high cost, because those produced in other cell lines or organisms have different patterns of glycosylation and subsequently a reduced in vivo activity. One alternative could be use a rhEPO without glycosylations, however, this variant has almost null in vivo activity, despite of having a higher in vitro activity and affinity for its receptor (EPOR) than native hEPO. The glycosylations generates limitations for the production of rhEPO, but they give important advantages to the protein like an augmented molecular weight and a decreased affinity for EPOR. In this review we summarized the principal biotechnological problems of the rhEPO glycosylations and we give insights of some possible solutions that have been tried to overcome those issues.