1. Absorbance changes were measured following stimulation of single muscle fibres injected with the metallochromic indicator dye Arsenazo III. Two dye-related signals can be clearly resolved: (1) an early, transient isotropic signal that appears to be due to the formation of Ca(2+):dye complex and (2) a slower, transient signal that is ;dichroic' in nature. The dichroic signal is obtained by taking the difference between absorbance changes measured with light plane polarized along the fibre axis (0 degrees light) and at right angles to the axis (90 degrees light).2. The time course of the dichroic signal is the same at all wavelengths employed, suggesting that a single underlying process is involved. The wavelength dependence of the magnitude of the signal is similar to that obtained for dye absorbance in a resting fibre.3. At 570 nm, near the isosbestic wavelength for changes in H(+):dye, Mg(2+):dye and Ca(2+):dye, the dichroic signal is near maximal. The absorbance change with 0 degrees light is positive and is about twice as large as the change with 90 degrees light, which is negative. This finding is consistent with the idea that the dichroic signal arises from dye molecules which change their orientation in the radially symmetric muscle fibre. The direction of the change is for the dye's transition moment to become more aligned with the fibre axis during activity.4. During a train of ten action potentials the (isotropic) Ca(2+) transient increases in magnitude three-fold, whereas the dichroic waveform reaches a plateau value only 30-40% larger than the single twitch value.5. Replacing H(2)O in Ringers with D(2)O causes a slight reduction in the Ca(2+) signal, reduces the dichroic signal to 0.4 times normal, and reduces tension to 0.1 times normal. Qualitatively similar reductions were found to accompany an increase in osmolality of H(2)O Ringer from 1 x to 2.5 x normal.6. Dichroic signals are also observed in fibres injected with Dichlorophosphonazo III. These are similar in many respects to the Arsenazo III dichroic signals.7. With Arsenazo III, the dichroic signal probably arises from a reorientation of some dye molecules which are bound to one of the oriented structures in muscle. The reorientation lags the Ca(2+) transient and may be due to a change which occurs in the oriented structure itself. Using this idea, the Arsenazo III dichroic signal can be fitted by assuming that Ca(2+) ions bind to receptor sites and that this binding induces the required change in the oriented structure. The analysis indicates that the hypothetical receptors have a dissociation constant for Ca(2+) equal to 0.1-1 times the peak value of myoplasmic free [Ca(2+)] during a twitch and an ;off' rate constant equal to 10-30 sec(-1) at 15 degrees C.