Potassium supplements have been recommended to hasten recovery and prevent cardiopulmonary complications in patients with thyrotoxic periodic paralysis (TPP). However, this recommendation has not yet been proven efficacious. Thirty-two patients with acute attacks of TPP over a 3-year-period were divided into 2 groups. Group A (n = 12) was a control group treated with normal saline infusion 125 mL/hr only. Group B (n = 20) received intravenous KCl administration at a rate of 10 mmol/hr in normal saline 125 mL/hr. During the attack and for 6 hours after muscle recovery, hemodynamics were continuously recorded and muscle strength and plasma K(+) concentration were measured hourly. The sex, age, muscle strength, thyroid function, biochemical values including plasma K(+) levels, as well as the time from attack to therapy (3.6 +/- 1.6 v 3.3 +/- 1.0 hr) were not significant between the 2 groups. However, recovery time was significantly shorter in the KCl group than the control (6.3 +/- 3.8 v 13.5 +/- 7.5 hr, P < .01). Rebound hyperkalemia greater than 5.5 mmol/L occurred in 40% patients receiving KCl. The dose of KCl administered and peak K(+) concentration were positively correlated (r = 0.85, P < .001). In conclusion, KCl therapy proves to help the recovery of paralysis in TPP associated with rebound hyperkalemia. KCl supplementation should be given as small as possible (<10 mmol/hr) to avoid rebound hyperkalemia unless there are cardiopulmonary complications.