Repolarisation alternans has been found to be predictive of or precede lethal ventricular arrhythmias in cardiac patients, as well as in the laboratory. Alternans is believed to cause arrhythmias by enhancing the regional physiological heterogeneities within normal cardiac tissue, to the point where functional conduction blocks develop, to initiate re-entrant arrhythmias. Using qPCR, immunohistochemistry and western blot (to confirm the specificity of antibodies), this study investigated the expression of ion channels and extracellular matrix (mRNAs and proteins) in areas of ventricular tissue displaying cardiac alternans & non-alternans sites. Epicardial samples were obtained from 19 patients who were undergoing elective surgery primarily for ischaemic heart disease. During the surgery, the patient’s heart was mapped using an epicardial sock which contained 256 electrodes. The heart was then paced at fixed cycle lengths to induce alternans. Activation Recovery Intervals (ARI), a surrogate for action potential duration (APD), was estimated and repolarisation alternans was quantified by an index incorporating both extent and persistence of beat-to-beat variations in ARI. Two myocardial biopsies were taken for mRNA analysis, each from sites of maximum and zero repolarisation alternans. The criterion for alternans was ≥ 6 consecutive beats of alternans with a steady state drive train of ventricular pacing at between 600-350 ms cycle length for 30 seconds each (50 ms reduction in cycle length between drive trains). Vimentin mRNA, which is a fibroblast marker, and collagen type 1 mRNA, which constitutes ~80% of total collagen mRNA, showed positive correlations with repolarisation alternans (R2 of 0.76 and 0.86 for vimetin and collagen type I, respectively). KChIP2 mRNA showed a positive correlation with repolarisation alternans (R2=0.82), whereas SERCA2 mRNA showed a negative correlation with repolarisation alternans (R2=0.82). A possible explanation is that there are areas of micro-infarction in the ventricle and this causes fibrosis accompanied by a downregulation in KChIP2 and SERCA2 which contribute to the production of repolarisation alternans. These micro-infarcts may, therefore, be responsible for arrythmogenesis.