We examine whether and the extent to which news-based sentiment, captured by textual analysis, can predict the performance of the private commercial real estate market in the United States. Our results show that sentiment reflected in news abstracts of The Wall Street Journal predicts returns of commercial real estate up to four quarters in advance. These findings are statistically significant and persist even when controlling for other related factors. This suggests that news-based sentiment can serve as an early market indicator. We are the first to examine the bidirectional relationship between sentiment, measured by textual analysis, and the performance of the private U.S. commercial real estate market. The findings contribute to the academic literature, and carry practical implications for real estate professionals.