Postmortem serum digoxin levels from any source routinely exceed antemortem values. Variation resulting from site of sampling gave a mean postmortem to antemortem ratio of 1.96 for heart, 1.63 for subclavian vein, and 1.42 for femoral vein samples. No correlation could be made between the postmortem interval and the increase in post-mortem serum values, irrespective of the site of sampling. A combination of femoral venous serum and vitreous humor values gave the best information for determining possible antemortem digoxin toxicity.