Cassava sweet potato intercropping experiments were carried out at Ukiriguru (semiarid zone), Tanzania during the period from 1989-91. Four cassava varieties used were, Msitu Zanzibar, Alpim valenca, Mzimbirala and Liongo control. The second experiment tested four cassava planting densities, 6666, 10000, 13333 and 20000 plants per hectare. In the two trials, cassava was planted with single or double rows of sweet potato on the same ridge. Basal dressing of inorganic fertilizers at the rate of 60 N, 30 P, and 30 K kg/ha was applied to one half while other half was left unfertilized. Results showed that cassava and sweet potato can be successfully and profitably intercropped since they have different growth durations that allowed for yield gains through better utilization of space and time. A sound fertilization practice in cassava/ sweet potato associations was deemed necessary to ensure soil fertility maintenance. Although nutrient use was a limiting factor for increased productivity, other potential agronomic practices such as the use of organic matter, crop residues, farmyard manure and integration of leguminous crops, to improve soil texture and nutrients may be additional attractive options for farmers. Some of the other aspects that need more research are, adjusting cutting length of cassava to avoid early shading of the intercrop, better rooting cultivar use, and type of seedbed used for sweet potato cuttings. Further studies are also warranted to know specific nutrient requirements in this intercrop and critical nutrient demand times of the individual crops in this association.