Electrophoresis, immunoblots, immunohistochemistry and image analysis methods were applied to characterise canine trunk and appendicular muscle fibres according to their myosin heavy chain (MyHC) composition and to determine, on a fibre-to-fibre basis, the correlation between contractile [MyHC (s), myofibrillar ATPase (mATPase) and sarco(endo)plasmic reticulum Ca(2+)-ATPase (SERCA) isoforms], metabolic [succinate dehydrogenase (SDH) and glycerol-3-phosphate dehydrogenase (GPDH) activities and glycogen and phospholamban (PLB) content] and morphological (cross-sectional area and capillary and nuclear densities) features of individual myofibres. An accurate delineation of MyHC-based fibre types was obtained with the developed immunohistochemical method, which showed high sensitivity and objectivity to delineate hybrid fibres with overwhelming dominance of one MyHC isoform. Phenotypic differences in contractile, metabolic and morphological properties seen between fibre types were related to MyHC content. All canine skeletal muscle fibre types had a relatively high histochemical SDH activity but significant differences existed in the order IIA>I>IIX. Mean GPDH was ranked according to fibre type such that I<IIA<IIX. Type IIA fibres were the smallest, type IIX fibres the largest and type I of intermediate size. Capillary and nuclear density decreased in the order IIA>I>IIX. Hybrid fibres, which represented nearly one third of the whole pool of skeletal muscle fibres analysed, had mean values intermediate between their respective pure phenotypes. Slow fibres expressed the slow SERCA isoform and PLB, whereas type II fibres expressed the fast SERCA isoform. Discrimination of myofibres according to their MyHC content was possible on the basis of their contractile, metabolic and morphological features. These intrafibre interrelationships suggest that myofibres of control dogs exhibit a high degree of co-ordination in their physiological, biochemical and morphological characteristics. This study demonstrates that canine skeletal muscle fibres have been misclassified in numerous previous studies and offers useful baseline data and new prospects for future work on muscle-fibre-typing in canine experimental studies.