This chapter considers a competing risks model for one-shot device testing data with multiple failure modes collected from constant-stress accelerated life-tests (CSALTs). The likelihood approach, via expectation-maximization (EM) algorithm and the Bayesian approach with various priors, is discussed for the estimation of model parameters. Two different scenarios are considered for the observed data: first in the case when the observed data contain no masking, and then the case when masking is present in the observed data. The chapter evaluates the performance of the EM algorithm for one-shot device testing data with competing risks under Weibull lifetime distribution, and considers the same CSALTs for exponential lifetime distribution. It also evaluates the performance of the EM algorithm and the Bayesian approach with various priors for one-shot device testing data with competing risks under exponential lifetime distribution.