The primary periodic paralyses are a group of very rare, genetic ion channel disorders characterized by transient attacks of severe flaccid weakness accompanied by serum potassium levels that are high, low, or normal. 1,2 In patients who present with characteristic episodes of flaccid weakness with abnormal potassium levels, especially when a family history is present, or in patients in whom genetic testing for one of the periodic paralyses confirms a pathogenic mutation, the clinical diagnosis is made rather easily. However, in those patients for whom this is not the case, the long exercise test (LET) is an electrodiagnostic technique that can be used to help establish the diagnosis. The LET is performed in the electromyography laboratory and entails measuring the baseline compound muscle action potential (CMAP) area or amplitude at rest and then measuring the CMAP amplitude or area over a period of 40-60 min after 5 min of isometric exercise of the recorded muscle, looking for an increment followed by a prolonged gradual decline. This methodology takes advantage of the fact that, in the periodic paralyses, attacks of weakness are known to be provoked by periods of rest after vigorous exercise. Although much has been written about the LET in the periodic paralyses, there are no firmly agreed upon cutoff values for amplitude and area decrements to define an abnormal LET result, and methods used to document these decrements vary. [3][4][5][6][7][8] In addition, cutoff values have classically been determined by studying a normal population, determining the mean and SD and setting the cutoff values at 2 SD above the mean.The study by Simmons et al. 9 in this issue of Muscle & Nerve elegantly summarizes the methodological issues and discrepancies that have been observed in previous reports of the LET in the primary periodic paralyses. They offer a logical, coherent set of LET cutoff guidelines that are based on their analysis of a cohort of 55 patients with genetically or clinically confirmed periodic paralysis and 125 control patients (healthy volunteers and patients with other neuromuscular disorders) by using a Bayesian analysis. Bayesian analysis, named for the English mathematician Thomas Bayes, is a statistical procedure that takes into account known information about a particular sample population variable that is being studied to guide and interpret the statistical analysis. 10 Simply put, the fundamental question when performing a diagnostic test is: how much does an abnormal test influence the likelihood that the patient truly has the disorder? If a clinician assesses the probability that an individual has the disorder as low (e.g., 25%) and a positive test result increases the probability to 35%, the test is of little use (despite being positive). However, if a positive test increases the probability to 95%, the diagnosis is almost certain. Bayesian analysis has been used previously to optimize the diagnosis of various entrapment neuropathies. 11 Bu using Bayesian analysis, the pretest probability distrib...