Additional lead electrocardiogram (ECG) has been proved to be useful in the diagnosis of acute myocardial infarction. However, the additional leads are not routinely employed due to limits of traditional electrocardiograph systems, and inconvenience and strains on patients during measurement. Our research team has developed a method to synthesize a virtual ECG on an additional lead based on computer simulation of heart. The synthesized additional leads ECG makes it possible that more detailed information can be obtained conveniently by using existing electrocardiograph. In this research, ECGs on standard 12 leads and additional leads are measured from 302 subjects. 150 subjects' data are randomly selected and used as learning data to estimate the transfer coefficient for deriving three right ventricular leads V3R, V4R, and V5R, and three posterior chest leads V7, V8, and V9 based on the method proposed by us. Then, the other subjects' ECG on additional leads are synthesized from the 12 lead system and compared with the actually measured counterparts. Difference of Q wave, R wave, and ST segment, and the correlation of averaged ECG are defined and estimated from the synthesized and measured data as evaluation indexes. Evaluation results demonstrate that virtual additional leads can be derived from 12 standard leads with a high accuracy qualified for clinical diagnosis of cardiac diseases. © 2012 IEEE.