Water vapor in the upper troposphere is one the largest uncertainties in NWP and climate research. Advanced microwave sounding unit (AMSU) on-board NOAA satellite series is capable of providing water vapor data even in the presence of non-precipitating clouds. The study presents the retrieval of three quantities, namely, water vapor, temperature, and surface emissivity from AMSU data based on optimal estimation method. The covariance matrices of temperature and water vapor are calculated using radiosonde observations. AMSU-A channels are used for temperature retrieval and AMSU-B channels are used for water vapor retrieval. The retrieval of water vapor and temperature is validated using high resolution radiosonde data. One of the challenges encountered while the retrieval was the uncertainty in determining microwave surface emissivity. For surface emissivity retrieval, three window channels at 23.8 GHz, 31.4 GHz, and 50.3 GHz are used. The retrieved emissivity values show diurnal and seasonal variations. The emissivity retrieval is validated in two ways. One way is to compare the retrieved values with results from other emissivity algorithm. Another way is to see whether simulated brightness temperature agree with the measured brightness temperature. The data from NOAA-15 and NOAA-16 are used in this study.