There is evidence of growing disparities in primary schooling rates between urban and rural areas in Tanzania. This paper presents empirical estimates for the determinants of primary school attendance in Tanzania for the early 1990s, and provides a comparison of attendance rates between the urban and rural areas for a number of different age groups. All the estimated models provide adequate fits to the data and many of the estimated coefficients are consistent with prior expectations. A statistically significant differential in primary school attendance rates between urban and rural areas is detected for the age groups examined. On the basis of our estimates, a large part of the differential is attributable to differences in observed characteristics with an important role exerted by urban-rural differences in the measure used to proxy household income.