Predicted sodium concentrations [Na(+)] based on traditional calculations for the correction of hyponatremia often do not match treated [Na(+)], for various reasons. In many situations, hyponatremia is corrected at unexpectedly rapid rates. The authors present an analysis of two cases of overly rapid correction of hyponatremia despite apparently appropriate management based on initial evaluations. The mistakes involved are discussed and simple calculations demonstrated to prove that the overcorrections did not occur at random. Overcorrection in one case involved miscommunications between the emergency room and admitting physicians regarding the amount of saline and potassium already administered to the patient. Unexpected hypoosmotic polyuria was responsible for overcorrection in the other case. Overcorrection of hyponatremia may be preventable in many cases. In general, overcorrection of hyponatremia is caused by either "too much salt (Na(+) + K(+)) gained" or "too much water lost." Recognizing common pitfalls will enable physicians to avoid overcorrection and its attendant risk of fatal osmotic demyelinating syndrome (ODS).