OBJECTIVE: This study tested the reliability and validity of four definitions of rapid cycling. METHOD: Two trained psychiatrists, using the Schedule for Affective Disorders and Schizophrenia, independently assessed 210 patients with bipolar disorder. They checked whether each patient met four definitions of rapid cycling: one consistent with DSM-IV criteria, one waiving criteria for duration of affective episodes, one waiving such criteria and requiring at least one switch from mania to depression or vice versa during the reference year, and one waiving duration criteria and requiring at least 8 weeks of fully symptomatic affective illness during the reference year. The interrater reliability was calculated by Cohen’s kappa statistic. Patients who met each definition according to both psychiatrists were compared to those who did not meet any definition (nonrapid-cycling group) on demographic and clinical variables. All patients were followed up for 1 year. RESULTS: Kappa values were 0.93, 0.73, 0.75, and 0.80, respectively, for the four definitions of rapid cycling. The groups meeting the second and third definitions included significantly more female and bipolar II patients than did the nonrapid-cycling group. Those two groups also had the lowest proportion of patients with a favorable lithium prophylaxis outcome and the highest stability of the rapid-cycling pattern on follow-up. The four groups of rapid-cycling patients did not differ significantly among themselves on any of the assessed variables. CONCLUSIONS: The expression "rapid cycling" encompasses a spectrum of conditions. The DSM-IV definition, although quite reliable, covers only part of this spectrum, and the conditions that are excluded are very typical in terms of key validators and are relatively stable over time.