In previous work we have shown that depressing Ca²⁺-induced Ca²⁺ release either directly by applying tetracaine (Díaz et al., 2002) or indirectly by decreasing the size of the activating L-type Ca²⁺ current (Díaz et al., 2004) results in a beat to beat alternation of the amplitude of the systolic Ca²⁺ transient. Under these conditions, the larger Ca²⁺ transients resulted from initial unsynchronized release of Ca²⁺ from the sarcoplasmic reticulum that then propagated as Ca²⁺ waves throughout the cell. Beat to beat alternation of the amplitude of cardiac contraction and the action potential duration is commonly seen in heart failure (Kodama et al., 2004). It is not, however, known whether unsynchronized Ca²⁺ release or Ca²⁺ waves are relevant under these conditions. In order to investigate this question we have used a model of heart failure. Under isoflurane inhalational anaesthesia, pacemakers were implanted in sheep and the ventricle was stimulated at 210 beats per minute. Heart failure was apparent after 4 weeks. Left ventricular midmyocardial myocytes were isolated and loaded with the fluorescent indicator fluo-3. [Ca²⁺]_i was measured using confocal linescans. Action potentials were stimulated by depolarizing current pulses at 0.5 Hz using the perforated patch technique at room temperature. In 5 out of 20 cells alternation of the action potential duration (measured at 90% repolarization) was seen with the longer action potentials being 111 ± 3 % of the shorter ones. Unsynchronized Ca²⁺ release and resulting Ca²⁺ waves were seen in all these 5 cells. The amplitude of the [Ca²⁺]_i alternated on a beat to beat basis with different regions of the same cell alternating out of phase. We conclude that alternans produced by heart failure shows similar changes of Ca²⁺ handling to those observed when Ca²⁺ induced Ca²⁺ release is depressed.