This study applied panel-based stationary tests to investigate the nonstationarity of long-run tourist arrivals to Taiwan. The sample of visitors in this study arrived from six main source countries between 1970 and 2011. We employed a panel KPSS test using a Fourier function and sharp drifts, which considers cross-sectional dependence as well as sudden impact and smooth transformation of data. This test can determine, with more power, whether tourist arrivals are stationary. Results of this study indicate that the trend in the series of tourist arrivals to Taiwan is stationary, with the exception of arrivals from Korea and Europe, implying that economic fluctuation inserts more impact on the tourism industry in these countries than others.