Transaction costs are important in determining investment performance and can substantially reduce or possibly outweigh the expected value created by an investment strategy. Despite the increasingly prominent role of transaction costs in both practice and research, estimates of transaction costs are not always available or, where available, subject to considerable expense and error.This research proposes a new method of obtaining estimates of transaction costs using Limited Dependent Variable (LDV) Model in Jakarta Stock Exchange (JSX) over the time period 1995-1996. The primary advantage of this model is that it requires only the time series of daily security return, market index return and foreign exchange return, making it relatively easy and inexpensive to obtain estimates of transaction cost for all firms for which daily security return are available.The premise of this model is that if the value of information signal is insufficient to exceed the costs of trading, then marginal investor will either reduce trading or not trade, causing zero return. This model of transaction costs is based on the number of zero returns. A security with high transaction costs will have less frequent price movements and more zero returns than a security with low transaction cost.This research finds that zero returns are very frequent. As much as 79,216% of the smallest market capitalization-based firm's daily security returns are zero during a year, and for the largest market capitalization-based firms, 17,682% of the daily security returns are zero. The proportions of zero returns are inversely related to market capitalization at 0,01 level of significance. These proportions are highly correlated (-0,9878) with market capitalization.The estimates of effective round-trip transaction costs obtained from the LDV model range from 1,1038% for smallest market capitalization-based firms to 12,317% for the Iargest market capitalization-based firms. The transaction costs are inversely related to market capitalization at 0,01 level of significance. These estimates have an -1 correlation coefficient with market capitalization. Based on these findings, I am assured that LDV model can be employed in estimating transaction costs in Jakarta Stock Exchange over the time period 1995-1996.