Patients may complain of pain at the injection site after subcutaneous (s.c.) administration of erythropoietin (EPO). Local pain due to s.c. EPO into the thigh was evaluated in 60 hemodialysis patients in a double-blind, placebo-controlled study. Identical volumes and concentrations (2000 IU in 0.5 ml) of phosphate-buffered epoetin-alpha (EPO-alpha ph), citrate-buffered epoetin-alpha (EPO-alpha ci) and epoetin-beta (EPO-beta) were compared to 0.5 ml of 0.9% saline (SAL), used as placebo. The patients received the 4 injections at the same occasion. For pain evaluation, a verbal scale ranging from no pain (0) to extremely painful (5) and a 10 cm ungraduated visual analogue score (VAS) (0 = no pain, 10 = maximal pain) were used. Treatment acceptance was assessed (yes/no) and expressed as a percentage of the population. Ranking of the preparations from 1 to 4 according to increasing local discomfort was performed. Median verbal pain scores and interquartile ranges were 1.0 (0-2) for SAL, 0.0 (0-2) for EPO-beta, 1.5 (0-3) for EPO-alpha ph (p < or = 0.05 vs SAL and EPO-beta) and 3.0 (2-4) for EPO-alpha ci (p < or = 0.001 vs EPO-alpha ph). VAS was 0.9 (0.5-2.5) for SAL, 0.9 (0.4-2.4) for EPO-beta, 2.7 (0.8-5.7) for EPO-alpha ph (p < or = 0.001 vs SAL and EPO-beta) and 4.2 (1.7-6.4) for EPO-alpha ci (p < or = 0.001 vs EPO-alpha ph). Treatment acceptance was 73% for SAL, 78% for EPO-beta, 60% for EPO-alpha ph (p < or = 0.05 vs EPO-beta) and 32% for EPO-alpha ci (p < or = 0.05 vs EPO-alpha ph). Ranking was 2 (1-3) for SAL, 2 (1-2) for EPO-beta, 3 (1-4) for EPO-alpha ph (p < or = 0.05 vs SAL and EPO-beta) and 4 (3-4) for EPO-alpha ci (p < or = 0.05 vs SAL and EPO-beta) and 4 (3-4) for EPO-alpha ci (p < or = 0.001 vs EPO-alpha ph). In conclusion, s.c. EPO-alpha ph is better accepted than s.c. EPO-alpha ci. However, s.c. EPO-beta is less painful.