The effect of serum-clot contact time on laboratory results was studied by dividing each blood specimen into four blood collection tubes. The control sera were separated from the clot within 30 min of the collection. The other tubes were incubated at 32 degrees C, and the sera were separated at 3, 6, and 24 h. The sera were stored at 4 degrees C and analyzed at the same time. The stability of the tests was determined by comparing the results of the 3-, 6-, and 24-h samples with the values from the 30-min samples. The acceptable limits around the 30-min values were derived from the analytical and intraindividual biological variation of the tests. A total of 63 analytes were studied. Potassium, phosphorous, and glucose were the least stable, and the serum should be separated from the clot within 3 h for these analytes. Albumin, bicarbonate, chloride, C-peptide, HDL-cholesterol, iron, LDL-cholesterol, and total protein should be separated within 6 h. The other analytes were stable for 24 h of serum-clot contact.