The purpose of this study is to develop the fragility curve associated with a real building. An approach to estimate seismic fragility without the need to perform the nonlinear dynamic analysis for determining the damage state of the structure as well as without the need to perform the Monte Carlo simulation for calculating the damage probability with respect to ground motion of different intensities is developed here. The capacity spectrum method is modified to determine the damage state of the target building. Consequently, nonlinear push over analysis is performed to obtain the capacity curve of the building and the damage states of the structure can then be determined by using the performance curve derived from capacity curve and code spectrum associated with ground motion within certain range of peak ground acceleration (PGA) values. Fragility curves in this study are assumed to be lognormal distribution described by two parameters and are developed as a function of PGA. The estimation of these two parameters is done by the maximum likelihood method treating each event of damage as a realization from a Bernoulli experiment. In fact, genetic algorithm (GA) is employed to obtain the parameters. The uncertainty associated with the ground motion is simulated by analyzing the statistical characteristics of the records of ground motion collected near the site. In addition, the uncertainties associated with the capacity spectrum and the thresholds of different damage states are also analyzed. Finally, fragility curves considering different level of uncertainties are constructed and compared for target building.