The interest in business-cycle asymmetry has beensteadily increasing over the past 15 years. Most research has focused onthe different behavior of macroeconomic variables during expansions andcontractions, which by now is well documented. Recent evidence suggests thatsuch a two-phase characterization of the business cycle might be toorestrictive. In particular, it might be worthwhile to decompose the recoveryphase in a high-growth phase (immediately following the trough of a cycle)and a subsequent moderate-growth phase. The issue ofmultiple regimes in the business cycle is addressed usingsmooth-transition autoregressive (STAR) models. A possiblelimitation of STAR models as they currently are used is that essentiallythey deal with only two regimes. We propose a generalization of the STARmodel such that more than two regimes can be accommodated.It is demonstratedthat the class of multiple-regime STAR (MRSTAR) models can beobtained from the two-regime model in a simple way.The main properties of the MRSTAR model and several issues that arerelevant for empirical specification are discussed in detail.In particular, a Lagrange multiplier-type test is derived that can beused to determine the appropriate number of regimes. A limitedsimulation study indicates its practical usefulness.Application of the new model class to U.S. real GNP provides evidence infavor of the existence of multiple business-cycle phases.