The thermal unfolding of myosin rod, light meromyosin (LMM), and myosin subfragment 2 (S-2) was studied by differential scanning calorimetry (DSC) over the pH range of 6.5-9.0 in 0.5M KCl and either 0.20 M sodium phosphate or 0.15M sodium pyrophosphate. Two rod samples were examined: one was purified by Sephadex G-200 without prior denaturation (native rod), and the other was purified by a cycle of denaturation-renaturation followed by Sephacryl S-200 chromatography (renatured rod). There were clearly distinguishable differences in the calorimetric behavior of these two samples. At pH 7.0 in phosphate the DSC curves of native rod were deconvoluted into six endothermic two-state transitions with melting temperatures in the range of 46-67 degrees C and a total enthalpy of 4346 kJ/mol. Under identical conditions the melting profile of LMM was resolved into five endothermic peaks with transition temperatures in the range of 45-66 degrees C, and the thermal profile of long S-2 was resolved into two endotherms, 46 and 57 degrees C. Transition 4 observed with native rod was present in the deconvoluted DSC curve for long S-2, but absent in the DSC curve for LMM. This transition was identified with the high-temperature transition detected with long S-2 and attributed to the melting of the coiled-coil alpha-helical segment of subfragment 2 (short S-2). The low-temperature transition of long S-2 was attributed to the unfolding of the hinge region. The smallest transition temperatures observed for all three fragments were 45-46 degrees C. It is suggested that the most unstable domain in rod (domain 1) responsible for the 46 degrees C transition includes both the hinge region, which is the C-terminal segment of long S-2, and a short N-terminal segment of LMM. This domain, accounting for 21% of the rod structure, contains the S-2/LMM junction, and upon proteolytic cleavage yields the C-terminal and N-terminal ends of long S-2 and LMM, respectively. Over the pH range of 6.5-7.5, the observed specific heat of denaturation of rod was approximately equal to the sum of the specific heats of LMM and S-2. This finding provides an additional argument for the existence of independent domains in myosin rod.