1. We measured intracellular Ca2+ transients during rapid cooling contractures (RCCs) in guinea-pig ventricular myocytes using the fluorescent Ca2+ indicator, Indo-1. 2. Rapid cooling of myocytes from 22 to 0-1 degrees C induced a rapid increase in [Ca2+]i which preceded the peak of the contraction and was sometimes large enough to saturate Indo-1. This indicates that [Ca2+]i may reach greater than 10 microM during an RCC. 3. The [Ca2+]i during the RCC slowly declined from its peak value and most of this decline in [Ca2+]i can be attributed to slow reaccumulation of Ca2+ by the sarcoplasmic reticulum (SR) in the cold. RCCs induced in the absence of Cao2+, were not different from control, supporting previous conclusions that RCCs depend exclusively on intracellular Ca2+ stores. 4. RCCs are depressed by long rest periods (rest decay) or by exposure to ryanodine or caffeine, which supports conclusions that RCCs are due to Ca2+ release from the SR. The rest decay of RCCs can be almost completely prevented by applying Nao(+)-free solution during the rest period. This implies that the loss of SR Ca2+ during rest depends on the sarcolemmal Na(+)-Ca2+ exchange (and not the sarcolemmal Ca2(+)-ATPase pump). 5. Rapid rewarming during an RCC normally leads to an additional transient contraction (or rewarming spike), without any increase in [Ca2+]i. Thus, the rewarming spike might be attributable to an increase in myofilament Ca2+ sensitivity induced by rewarming. 6. A second RCC is used to assess the fraction of Ca2+ which is re-sequestered by the SR during relaxation from the first RCC. In control solution progressive RCCs decline in amplitude, but in Na(+)-free, Ca2(+)-free solution they are of constant amplitude. We conclude that the SR Ca2+ pump and Na(+)-Ca2+ exchange are responsible for relaxation and that the latter may account for 20-50% of relaxation. 7. These results support the use of RCCs as a useful means of assessing SR Ca2+ content in intact cardiac muscle cells.