A new Enzyme Linked Immuno Sorbent Assay (ELISA) kit for the determination of serum ferritin has been compared with another ferritin kit based on the Immuno Radio-Metric Assay (IRMA) approach, both assays containing similar antibodies. Based on these studies, we found the within-run precision of the ELISA (and IRMA) to have coefficients of variation of 4-10% and 2-6% respectively, over a concentration range of 12-600 micrograms/l. The between-run precision for the same concentration range exhibited a CV range of 9-13% and 7-11% respectively. The sensitivities were found to be 1.4 micrograms/l and 0.9 microgram/l. The mean recovery was 103% for the ELISA procedure. It was found that, using the serum dilution technique, the linearity reached to 1000 micrograms/l. In the ELISA procedure no influence from the so-called "high dose hook effect" was observed. While EDTA-plasma produced 6% lower values than serum in the ELISA technique, no interference from albumin, gamma-globulins and mild haemolysis was observed. Stability problems with the ELISA kit were not encountered. A comparative analysis of multiple specimens demonstrated nearly identical values with r = 0.994 and y = 0.87 x1.01. The quality and ease of operation of the ELISA approach compared with other techniques are discussed. In conclusion it is possible to replace a radio-label in an immunoassay with an enzyme-label with the same degree of reliability and other parameters of quality control exhibited by radioimmunoassays.