Using a nested multinomial logit model (NMNL), this paper estimates the choice probabilities and demand elasticities of various modal alternatives for Mumbai using 1990 trip origination and destination data. The results indicate that with an increase in the cost of private transit system, market share is reduced. The result of this experiment is highly elastic private commuting mode demand. In addition, reduction in the market share of own vehicles is larger than that of hired vehicles. The increase in the cost of mass transit system reduces its share but is inelastic. Decrease in the market share of one mode is accompanied by increase in the market share of others. Income has a positive effect on the market share of private transit modes, particularly on the own vehicles' share. Other explanatory variables such as distance have a negative effect on mass transit system and duration has a negative elasticity for all modes of transportation.