It has been previously recognized that a number of protocols may cause breakage of the triad junction and separation of the constituent organelles of skeletal muscle. We now describe a fraction of triad junctions which is refractory to the known protocols for disruption. Triads were passed through a French press and the dissociated organelles were separated on a sucrose density gradient, which was assayed for PN200-110, ouabain and ryanodine binding. Ryanodine binding showed a single peak at the density of heavy terminal cisternae. On the other hand, the PN200-110 and ouabain, which are external membrane ligands, bound in two peaks: one at the free transverse tubule region and the other at the light terminal cisternae. Similarly, a two peak pattern of PN200-110 and ouabain binding was observed when triad junctions were broken by the Ca2+-dependent protease, calpain, which selectively hydrolyzes the junctional foot protein. The light terminal cisternae vesicles were subjected to three different procedures of junctional breakage: French press, hypertonic salt treatment, and protease digestion using calpain or trypsin. The treated membranes were then centrifuged on density gradients. Only extensive trypsin digestion caused a partial shift of ouabain activity into the free transverse tubule region. These observations suggest that the triads are a composite mixture of breakage susceptible, "weak," and breakage resistant, "strong," triads. Scatchard analysis of PN200-110 suggests that the transverse tubules of strong triads contain a relatively high number of dihydropyridine receptors compared to those of weak triads. Thin section electron microscopic images of the strong triads comparable to those of intact muscle are presented. © 1990 Springer-Verlag New York Inc.