Background and Objectives Erythropoietin response to anaemia has been reported to be suboptimal in HIV infected anaemic patients. This was a cross sectional comparative study conducted to estimate the serum erythropoietin levels of anaemic HIV infected subjects in Lagos. The study aimed to assess the relationship between erythropoietin levels and degree of anaemia, CD4+ cell count and viral load of HIV infected treatment naive subjects and to ascertain the association with some inflammatory cytokines (such as CRP, TNF-α, IL-6) and ferritin. Subjects and Methods The study population consisted of a total of 120 subjects aged 18-66 years made up of four categories of individuals which included: (1) HIV infected treatment naive subjects who had anaemia with Hb level <10g/dL. Control subjects (who were age and sex matched) were: (2) HIV infected non-anaemic, (3) non-HIV infected anaemic, and (4) non-HIV infected non-anaemic individuals. A brief general clinical history and assessment was carried out and questionnaires were administered to each subject. Full blood count using automated counter- Sysmex 21N; CD4+ cell count by semi-automated cytometry- Partec flow cytometer; and viral load using Roche amplicon were estimated. Serum erythropoietin levels using ELISA kit supplied by ALPCO diagnostics 26G Keewaydin Drive Salem, NH 03079; serum ferritin levels; CRP levels; TNF- α levels; and IL-6 levels were estimated using ELISA kit and technique supplied by Assaypro (www.assaypro.com). Appropriateness of EPO response was determined by calculating EPO response ratio given by value of measured EPO levels as numerator and expected EPO level as denominator. The expected EPO level was deduced from computer generated regressional equation when EPO levels of HIV negative anaemic control subjects was plotted against their haemoglobin values. Ratio less than 0.8 was accepted as suboptimal response.