In this study we examined the impact on soil C, total soil N and available P of six rotations namely: long season tobacco cultivar ‘KM10’ grown continuously (ContKM10), medium season tobacco cultivar ‘RK8’ grown continuously (ContRK8), grass-grass-grass-KM10 (G-G-G-KM10), grass-grass-grass-RK8 (G-G-G-RK8), KM10-Crotalaria juncea (KM10-Cr) and RK8-Crotalaria juncea (RK8-Cr). The experiment was established in 1990 under irrigation on a sandy loam soil at Kutsaga Research Station, Zimbabwe. Soil samples were taken from 0- to 15-cm deep, after each season. After 9 years, tobacco-grass rotations showed higher soil C than monocropping, regardless of variety. The monocropping systems, ContKM10 and ContRK8, did not differ from KM10-Cr and RK8-Cr respectively showing that when crop intensity is maintained soil C will be reduced regardless of a winter C. juncea green manure in a sandy loam soil. After 9 years, soil N was greatest in the G-G-G-KM10 rotation. Available P was lower in the grass (G-G-G-KM10, G-G-G-RK8) relative to the other rotations regardless of variety. Available P accumulated in monocropping systems (ContRK8, ContKM10) and was consistently lower in the grass-tobacco rotations. This indicated an accumulation of P in the case of monocropping systems because of continuous inorganic fertiliser input. The results reaffirmed the deleterious effect of monocropping and suggested the need for diverse rotations.