tslstarmod performs an estimation of a logistic smooth transition autoregressive regression (LSTAR) model for time series data. This command allows estimating an endogenous structural break point in a time series data. The endogenous threshold when found is determined smoothly, contrarily to brutal transitions. In this regard, the LSTAR model can be considered as a generalization of the usual autoregressive process because the transition function is a smooth logistic function. The command allows to test the presence of an LSTAR model against a presence of a linear autoregressive model. It also handles the determination of the delay parameter. The theory behind the command tslstarmod can be found, for instance, in Terasvirta (2004) and Enders (2015).