Digoxin-like immunoreactive substance (DLIS) was measured in 34 samples obtained from subjects not receiving digoxin: 10 uremic, 10 third trimester pregnancy, seven cord blood and seven normal. DLIS concentration was measured by four commercial polyclonal radioimmunoassay (RIA) systems: Clinical Assay (CA), Corning Immophase (CI), Diagnostic Products Double Antibody (DP), Kallestad Quantitope (KK), a monoclonal antibody (MA) assay and a Fluorescence Polarization Immunoassay (FPIA). In general, the cord blood samples were richer in DLIS. Digoxin immunoassays, MA and DP showed minimal interference by DLIS in all samples, whereas FPIA and CA exhibited the maximal cross-reactivity with DLIS. In cord blood samples, mean +/- SD DLIS concentration ranged from 0.41 +/- 0.13 (by CA) to 0.034 +/- 0.02 ng ml-1 as measured by MA assay. In uremics, the mean DLIS concentration was below the detection limit of all RIA assays. The FPIA method showed a higher degree of cross-reactivity to DLIS, especially in the cord and pregnancy samples (0.42 +/- 0.13 and 0.4 +/- 0.14 ng ml-1, respectively). DLIS in uremics was below the FPIA detection limit of 0.2 ng ml-1. Overall, the degree of interference by DLIS in decreasing order was FPIA greater than CA greater than CI greater than or equal to KQ greater than DP greater than or equal to MA. The cord blood samples were re-analysed by FPIA (Digoxin-II assay) 4 months later, resulting in 2-4-fold higher DLIS concentrations for these samples. This appears to be due to the substitution of 5-sulphosalicylic acid as a protein precipitating reagent and this effect may have been accentuated by freeze-thaw cycles.