In September 2005, the Chilean Competition Authority filed a lawsuit against five private health insurance providers for violation of antitrust laws. They were accused of colluding to reduce the coverage levels of their plans. Between March 2002 and March 2003, they reduced the coverage offered from 100 percent for hospitalization and 80 percent for ambulatory care to 90 percent and 70 percent, respectively. These facts were undisputed, but, logically, the mere observation of parallel conduct is not enough to infer collusion. In this article, we analyze the merits of the accusation and find strong support for it. We first present evidence on the definition of the relevant market (that excludes the public insurer) and the presence of entry barriers. We then present a simple model of imperfect competition that gives some testable predictions to separate the hypotheses that insurers colluded to reduce coverage and the one-argued during the trial by the insurers-that the coverage reduction was a competitive reaction to a cost shock. We then present econometric evidence consistent only with the hypothesis of collusion: The accused insurers reduced sales efforts and marketing expenses relative to the non-accused insurers, the monthly transfer rates of insurees among accused insurers decreased during the period of the alleged agreement relative to the rate of the non-accused insurers, and profits of both accused and non-accused insurers increased after the coverage reduction.