The aim of this study was to evaluate two enzyme immunoassays for the estimation of serum and salivary digoxin. The methods under review were the Enzyme Multiplied Immunoassay Technique and Cloned Enzyme Donor Immunoassay assays. Within and between assay precision for both immunoassays was good giving co-efficient of variations of less than 6.1%. Analytical recoveries ranged from 96-103%. Sensitivity was 0.28 nmol/1 of digoxin, approximately. Results from serum specimens from 52 patients using both methods and a RIA method from an external laboratory (Beaumont Hospital) were compared. A correlation of 1.008 was obtained for the EMIT and RIA methods, while 0.985 resulted from the EMIT and CEDIA methods. The RIA and CEDIA methods gave a correlation of 0.94. Haemoglobin, bilirubin, lipaemia, and matrix effects, i.e. (protein concentration) did not interfere with the estimation of digoxin by the EMIT method. In the CEDIA method, high concentrations of protein resulted in falsely low digoxin concentrations, whereas low protein concentrations gave falsely high digoxin concentrations. Haemoglobin, bilirubin, and lipaemia caused interferences with the assay which was concentration dependent. Digoxin-like Immunoreactive Factors were greatly reduced or eliminated by the EMIT technique but interfered with the CEDIA method resulting in "apparent digoxin concentrations" as high as 1.18 nmol/1. A further study was performed to investigate the possible use of the EMIT and CEDIA methods for salivary digoxin estimations. Accuracy, precision, sensitivity and linearity were found to be comparable with that of the serum digoxin methods. No procedural modification was required for the EMIT method, while minor modifications were needed for the CEDIA. The mean digoxin saliva/serum concentration ratio in 20 hospital patients using the EMIT method was 0.67 and the correlation was 0.96. The mean ratio using the CEDIA method was 0.62 with a correlation of 0.93. These results were obtained when serum and saliva samples were taken simultaneously. In each method, DLIF concentration less than the sensitivity range was obtained in saliva from patients in renal and hepatic failure and third trimester pregnancy. On the basis of all the factors assessed, the EMIT is the most suitable for routine use in the clinical chemistry laboratory.