Abstract: A field plot experiment was conducted to investigate the tobacco yield and different forms of soil phosphorus under tobacco garlic crop rotation and intercropping patterns. The results showed that compared with tobacco monoculture, the tobacco yield and proportion of middle/high class of tobacco leaves to total leaves were significantly increased in tobacco garlic crop rotation and intercropping, and the rhizosphere soil available phosphorus contents were 1.3 and 1.7 times as high as that of tobacco monoculture at mature stage of lower leaf. For the inorganic phosphorus in rhizos phere and non-rhizosphere soil in different treatments, the contents of O-P and Fe-P were the highest, followed by Ca₂-P and Al-P, and Ca₈-P and Ca₁₀-P were the lowest. Compared with tobacco monoculture and tobacco garlic crop intercropping, the Ca₂-P concentration in rhizosphere soil under tobacco garlic crop rotation at mature stage of upper leaf, the Ca₈-P concentration at mature stage of lower leaf, and the Ca₁₀-P concentration at mature stage of middle leaf were lowest. The Al-P concentrations under tobacco garlic crop rotation and intercropping were 1.6 and 1.9 times, and 1.2 and 1.9 times as much as that under tobacco monoculture in rhizosphere soil at mature stages of lower leaf and middle leaf, respectively. The OP concentrations in rhizosphere soil under tobacco garlic crop rotation and intercropping were significantly lower than that under tobacco monoculture. Compared with tobacco garlic crop intercropping, the tobacco garlic crop rotation could better improve tobacco yield and the proportion of high and middle class leaf by activating O-P, Ca₁₀-P and resistant organic phosphorus in soil.