Various methods for extraction and isolation of myocardial glycogen show different yields and identify different glycogen subsets. The aim of the present study was to identify a glycogen fraction exposed to changes during myocardial ischaemia. Endomyocardial biopsies from 10 pigs were sampled before cardioplegia, after cardioplegic arrest, and after reperfusion. Glycogen yields were compared following five extraction procedures: (1) hot alkaline tissue digestion, (2) homogenization in perchloric acid and subsequent determination in homogenate, (3) homogenization in perchloric acid and subsequent determination in supernatant, (4) homogenization in perchloric acid and subsequent determination in the precipitate redissolved in hot alkaline and (5) homogenization in homogenisation buffer with lysating capacity. Glycogen was isolated on filter-paper and determined enzymatically. Hot alkaline tissue digestion yielded the highest glycogen amounts (63.5 +/- 18.3 nmol/mg wet weight). Glycogen yields in perchloric homogenate and supernatant were 51%, perchloric precipitate 47%, and buffer 30% of these obtained with hot alkaline. Glycogen yields in hot alkaline were comparable to the sum of those obtained in perchloric supernatant ("acid extractable glycogen") and redissolved precipitate ("heavily extracted glycogen") confirming that glycogen yields obtained with hot alkaline digestion represent "total glycogen". Acid extractable glycogen showed superior analytical characteristics compared with the other methods. Acid extractable glycogen demonstrated a consistent decrease during ischaemia whereas total glycogen and glycogen extracted in homogenization buffer tended to decrease. Glycogen in perchloric precipitate remained unchanged during ischaemia. These findings support a revival of the concept that tissue contains two forms of glycogen. Decreases in myocardial glycogen content during myocardial ischaemia are best observed with acid extractable glycogen.