A serne residue located in the active site of myosin head (S1) was labelled by 9-anthroylnitrile, an amino group located in the central domain of S1 was labelled by 7-diethylamino-3-(4′-isothio-cyanato-phenyl)-4-methylcoumarin, a cysteine residue located near the C-terminus of S1 was labelled by 5-[2-((iodoacetyl)-amino)ethyl]-amino-naphthalene-1-sulfonic acid (1,5-IAEDANS) and a cysteine residue located near the C-terminus of the alkali light chain 1 was labelled with iodoacetamido-tetramethyl-rhodamine. Polarization of fluorescence of S1 was measured in solution (where it indicated the mobility of actin-bound S1) and in myofibrils (where it indicated orientation of probes) to check whether the anisotropy of S1 labelled at different positions depended on the molar ratio S1:actin. In solution, when increasing amounts of actin were added to a fixed amount of labelled S1 (i.e. when myosin heads were initially in excess over actin), anisotropy saturated at 1 mol of S1 per 1 mol of actin. When increasing amounts of S1 were added to a fixed amount of F-actin (i.e. when actin was initially in excess over S1), the anisotropy saturated at 1 mol of S1 per 2 mols of actin. In myofibrils, orientation of S1 was different when S1 was added at nanomolar concentration (intrinsic actin was in excess over extrinsic S1) then when it was added at micromolar concentration (excess of S1 over actin). The fact that the anisotropy of S1 labelled at different positions depended on the molar ratio excluded the possibility that changes were confined to one part of the cross-bridge and supports our earlier proposal that the two rigor complexes which S1 can form with F-actin differ globally in conformation.