Chloride fluxes are the main contributors to the resting conductance of mammalian skeletal muscle fibers. ClC-1, the most abundant chloride channel isoform in this preparation, is believed to be responsible for this conductance. However, the actual distribution of ClC-1 channels between the surface and transverse tubular system (TTS) membranes has not been assessed in intact muscle fibers. To investigate this issue, we voltageclamped enzymatically dissociated short fibers using a two-microelectrode configuration and simultaneously recorded chloride currents (ICl) and di-8-ANEPPS fluorescence signals to assess membrane potential changes in the TTS. Experiments were conducted in conditions that blocked all but the chloride conductance. Fibers were equilibrated with 40 or 70 mM intracellular chloride to enhance the magnitude of inward ICl, and the specific ClC-1 blocker 9-ACA was used to eliminate these currents whenever necessary. Voltage-dependent di-8-ANEPPS signals and ICl acquired before (control) and after the addition of 9-ACA were comparatively assessed. Early after the onset of stimulus pulses, di-8-ANEPPS signals under control conditions were smaller than those recorded in the presence of 9-ACA. We defined as attenuation the normalized time-dependent difference between these signals. Attenuation was discovered to be ICl dependent since its magnitude varied in close correlation with the amplitude and time course of ICl. While the properties of ICl, and those of the attenuation seen in optical records, could be simultaneously predicted by model simulations when the chloride permeability (PCl) at the surface and TTS membranes were approximately equal, the model failed to explain the optical data if PCl was precluded from the TTS membranes. Since the ratio between the areas of TTS membranes and the sarcolemma is large in mammalian muscle fibers, our results demonstrate that a significant fraction of the experimentally recorded ICl arises from TTS contributions.