AbstractCompeting risks arise when a subject is exposed to many causes of failure. Data consist of the time the subject failed and an indicator of which risk caused the subject to fail. Examples in medicine include the analysis of cause to death data, the analysis of relapse and death in remission in cancer studies, or random right censoring. In engineering applications competing risks arise when analyzing series systems. Classical competing risks deal with the modeling of the probability of failure in the observed system (crude probabilities) or in systems with some causes of failure removed (net or partial crude probabilities). Modern competing risks theory, often based on counting process methodology, examines estimation techniques for competing risks probabilities, regression modeling for competing risks probabilities and comparing competing risks probabilities between treatment groups. Copyright © 2010 John Wiley & Sons, Inc.This article is categorized under:Statistical and Graphical Methods of Data Analysis > Reliability, Survivability, and Quality Control