This survey is devoted to the modelling and the estimation of reduced-form transition models, which have been extensively used and estimated in labor microeconometrics. The first section contains a general presentation of the statistical modelling of such processes using continuous-time (event-history) data. It also presents parametric and nonparametric estimation procedures, and focuses on the treatment of unobserved heterogeneity. The second section deals with the estimation of markovian processes using discrete-time panel observations. Here the main question is whether the discrete-time panel observation of a transition process is generated by a continuous-time homogeneous Markov process. After discussing this problem, we present maximum-likelihood and bayesian procedures for estimating the transition intensity matrix governing the process evolution. Particular attention is paid to the estimation of the continuous-time mover-stayer model, which is the more elementary model of mixed Markov chains.