Sartorius muscle fibres from cold-adapted Rana temporaria were exposed to variants of an established detubulation procedure (Koutsis et al. (1995) J Muscle Res Cell Motil 16, 519-528) to test the extent to which detubulation and tubular vacuolation phenomena could be separated using different conditions of osmotic shock. A control procedure was optimised to a 28-min exposure to 400 mM glycerol-Ringer. This was followed by a recovery step involving its replacement by a Ca2+/Mg(2+)-Ringer solution and steady cooling over 30 min from room temperature (approximately 18 degress C) to approximately 10 degress C, followed by the restoration of the normal Ringer solution. This procedure successfully abolished the action potential after-depolarisation component, reflecting a loss of tubular conduction ('detubulation') in 74.3 +/- 5.9% of the fibres studied. Omitting the cooling during the recovery step sharply reduced the incidence of detubulation. So did omitting either the high-[Ca2+] and/or [Mg2+] in the recovery solutions in test procedures, but to significantly different extents (P < 5%). Yet trapping of fluorescent Sulfhorhodamine B dye in 'closed' vacuoles persisted albeit with reduced proportions of fibre volume occupied by vacuoles. Furthermore, the variations in recovery conditions produced similar levels of vacuolation despite smaller vacuole sizes in the cooled fibres (P < 0.05). These findings demonstrate that fibre vacuolation and detubulation are phenomena that are potentially separable through varying the conditions of osmotic shock, with detubulation requiring significantly more stringent conditions than vacuolation.