The influence of short-term changes in stimulation pattern, both on the strength of contraction and on the amplitude of intracellular free-Ca2+ transients, was investigated in ferret papillary muscles. Intracellular free-Ca2+ concentration ([Ca2+]) was assessed from the luminescence emitted from muscles microinjected with the Ca2+-sensitive photoprotein aequorin. The relationships between the strength of contraction and changes in stimulation pattern lasting 1-2 beats could be described by monoexponential functions, all with very similar time constants (approximately 750 ms at 30 degrees C). Over the entire range that could be obtained, the strength of contraction, quantified by either peak tension or peak rate of tension development, was found to be linearly correlated with peak estimated [Ca2+]. Potential errors in the estimation of [Ca2+] from aequorin luminescence were analysed. To assess the influence of spatial non-homogeneities of [Ca2+] on the estimate of [Ca2+], a model of Ca2+ diffusion in heart muscle was developed. The possible effect of using an inaccurate calibration curve was also examined. The results of these analyses indicate that [Ca2+] estimated from aequorin luminescence should be proportional to, if not equal to, true spatial average [Ca2+] (errors less than 7%). Given the conclusion of the analysis described above, it is inferred from points 2 and 3 that the relationships between peak spatial average [Ca2+] and short-term changes in stimulation pattern are also represented by monoexponential functions, with time constants closely similar to those for the mechanical measurements. Exposure to ryanodine, a substance believed to inhibit the release of Ca2+ from sarcoplasmic reticulum, produced striking alterations in the pattern of variations in [Ca2+] mentioned above. These alterations are consistent with the hypothesis that the functions described above depend essentially upon properties of the sarcoplasmic reticulum.