It is not uncommon for patients to present to the emergency room with severe weakness and a markedly low plasma potassium concentration. We attempted to identify useful clues to the diagnosis of hypokalaemic periodic paralysis (HPP), because its acute treatment aims are unique. We retrospectively reviewed charts over a 10-year period: HPP was the initial diagnosis in 97 patients. Mean patient age was 29+/-1.1 and the male:female ratio was 77:20. When the final diagnosis was HPP (n=73), the acid-base state was normal, the urine K(+) concentration was low, and the transtubular K(+) concentration gradient (TTKG) was <3. In patients with thyrotoxic periodic paralysis (TPP) (n=39), hypokalaemia was very commonly accompanied by hypophosphataemia (1.9+/-0.1 mg/dl). A clinical diagnosis of sporadic periodic paralysis (SPP) was made if hyperthyroidism and a family history of HPP were both absent (n=29). One subgroup of patients with HPP had a severe degree of hypernatraemia (167+/-5.0 mmol/l, n=3). There were only two patients with familial periodic paralysis (FPP). In 24 patients, the initial diagnosis was HPP, but subsequent studies failed to confirm this diagnosis. Each of these patients had an acid-base disorder, a high rate of renal K(+) excretion in the presence of hypokalaemia, and a TTKG of close to 7. With respect to therapy, much less K(+) was given to patients with HPP, yet 1:3 subsequently had a plasma K(+) concentration that eventually exceeded 5.0 mmol/l. Using plasma acid-base status, phosphate and K(+) excretion parameters allows a presumptive diagnosis of HPP with more confidence in the emergency room.