This paper exploits a Markov regime-switching model to study the business cycle of the hotel industry (BCHI) in Taiwan from January 1999 to September 2011. The aggregate hotel room sales earnings are used to examine the business cycle of the Taiwanese hotel industry. The Markov regime-switching model detects two distinct regimes of the BCHI: a high-growth regime (HGR) and a low-growth regime (LGR). The average growth rate of HGR (LGR) is 4.12% (0.35%). The corresponding standard deviation in the two regimes is 5.60% and 0.30%, implying that LGR is more stable than HGR. The probability of the hotel industry staying in LGR (HGR) is 97% (66%) and the expected duration of HGR (LGR) is about 3 (39) months. Moreover, this study shows that the tourism market growth is significant in causing the hotel industry to stay in the HGR. The empirical findings provide useful information and policy implications for government tourism policymakers and tourist hotel managers.