This article presents a free-space terahertz time-domain spectroscopy (THz-TDS) system for nondestructive testing for moisture content and impurities in tobacco. The THz-TDS system is driven by two compact Ti:sapphire lasers with pulse widths of 20 and 45 fs, and enables rapid switching of laser sources. Analysis reveals a strong linear correlation between the THz signal amplitude and the moisture content of tobacco leaves, and distinct waveform differences are observed between tobacco leaves and impurities. Machine learning algorithms realize 99.9% accurate impurity classification and rapid moisture prediction with a mean absolute error of 0.28%. The measurement and analysis process can be completed within 2 s.