Congestive heart failure (HF) is a major focus of medical research. Its incidence has greatly increased in recent decades because of an aging population base and the increasingly successful treatment of other forms of chronic cardiac disease. Relevant large animal models of HF should reflect the complex interactions of cardiac dysfunction, neurohumoral dynamics and peripheral vascular abnormalities found in human HF. A number of large animal models have been developed, especially in dogs, sheep and swine, using naturally occurring HF, or single or combinations of interventions, as instruments to trigger the development of HF. Naturally occurring HF models are not commonly used because of ethical or perceived ethical grounds, however, King Charles Cavalier Spaniel and Yucatan Mini Pig models have been described. Tachycardia induced HF is the most commonly used HF model. Ventricular pacing at 220-240 bpm results in profound low output, biventricular, oedematous failure in two to three weeks. Lower pacing rates result in a more stable, sustainable, lesser degree of failure. Positive features of this model include 'acceptance', aetiological relevance to patient tachycardia induced HF, neurohumoral and functional profile similar to most human HF, relatively low cost simple preparation, ability to manipulate the degree of failure with pacing rate, reversibility, reliability and a large amount of published multi species data. Limitations to the use of the model are the rapid onset, the fact that reversibility is only relevant to the tachycardia induced patient HF, the absence of hypertrophy in failure, the diminished plasma atrial natriuretic peptide (ANP) levels, absence of ANP of ventricular origin, and the interference between rapid pacing and therapeutic interventions. Myocardial damage models of HF include those models induced by ischaemia, eg due to coronary occlusion (ligation or aneroid) or intracoronary microembolism, transmyocardial DC shock, toxic cardiomyopathy from adriamycin, doxorubicin or catecholamines. Overload models of HF may be induced by high pressure from aortic constriction, aortic regurgitation, renal artery constriction, pulmonary stenosis or aortocaval shunts, or by induction of mitral regurgitation from chordae or leaflet damage. No single, all-encompassing, large animal model of HF exists to date. Selection of the type of model to be used should be based primarily on the hypotheses to be tested and secondarily on the available resources and facilities.