Carsharing is considered an important alternative to the private car. Based on the studies in Europe and North America, it is agreed that carsharing can reduce private car ownership. However, due to the differences in the development stages and characteristics of cities and transportation, it is unknown whether carsharing can also decrease car ownership in China, which results in a gap in the evaluation of carsharing in China. This paper takes the carsharing project EVCARD in Shanghai as an example. Based on survey data on vehicle ownership from users and actual operations data, the impacts of carsharing on selling cars, giving up and delaying purchasing a car in Shanghai are quantitatively analyzed. The results in Shanghai are compared with the relevant research results in different cities around the world. The results show that: (1) the carsharing project in Shanghai also has the benefit of reducing private car ownership. Each carsharing vehicle can replace 4.56 private cars. However, unlike in European and North American cities, Shanghai carsharing mainly suppresses vehicle growth by reducing individual willingness to buy a car, because the proportion of cars sold owing to carsharing is obviously lower than in Europe and North America. (2) For giving up car purchases, carsharing mainly suppresses the first car in the family. (3) Good public transport and carsharing services can substantially improve the substitution benefits of carsharing for private cars. If carsharing can reach the service level of “whenever and wherever carsharing is available”, the proportion of users who sell, give up and delay buying cars will increase from 0.78%, 30% and 31% to 12%, 40% and 47% respectively. The findings in the study provide a quantitative basis for assessing the carsharing benefits on the reduction of vehicle ownership in China, and also provide a useful reference for evaluating carsharing benefits in other developing countries.