An evaluation of duplicate prothrombin time (PT) and activated partial thromboplastin time (aPTT) assays determined by the MCA 110 coagulation analyzer was undertaken to develop analytical duplicate performance criteria to quantitate the risks associated with single versus duplicate procedures. Included in the study were 1,277 patient samples. On the basis of the currently recommended therapeutic range for prothrombin ratios, a variation of approximately 10% or more between duplicates was considered to be unacceptable. For aPTT assays, the recommended therapeutic range for heparin therapy was usually 1.5 to 2.5 times the baseline value, and variations of up to 25% might be considered acceptable. With these relatively lenient criteria, approximately 2% of PT and 1.3% of aPTT assays had differences between duplicate values that were unacceptable. From this data the authors concluded that the frequency of errors produced by single estimations was too great for satisfactory clinical practice.