This report presents a summary of the meteorological data collected by the meteorological tower network in the Gregoire Lake region of northeast Alberta during the spring and summer of 1979. The network was established to help monitor the impact on the environment by the Amoco Canada Co. Ltd. pilot plant. A previous report, prepared by Athabasca Research Corporation (Ferguson 1979), presented an analysis of the winter 1978-79 data. A discussion is presented of the theoretical meteorological background including synoptic and mesoscale influences on the dispersion of effluents emitted into the atmosphere. Field dispersion experiments in the oil sands area are reviewed briefly. The statistics of the various weather elements are discussed. Wind velocity was measured at the 30 m tower level at Anzac, the Gregoire Lake Provincial Park, the Amoco pilot plant, and Stoney Mountain. Temperature, relative humidity, vertical velocity, precipitation, barometric pressure, and solar radiation were to be monitored at the pilot plant. The system for recording these data was not completely debugged by the end of the summer so data are not available yet. The meteorological statistics and the case studies indicated that when the air is stable or neutral, the regional airflow is deflected to follow the contours of the ridge, which is south of Gregoire Lake, and parallel to the Athabasca and Clearwater rivers. Under convectively unstable conditions, air flow tended to be upslope at the plant site and on Stoney Mountain, but similar to the regional flow at the valley stations. Wind speeds were generally light in agreement with long-term records in the oil sands area. Temperatures also followed the longterm trends. Examination of local meteorological and upper air data from Edmonton and Fort Smith indicated that the high concentrations of nitric oxide recorded on 27 September 1979 may have been related to low mixing heights. The source was probably not the pilot plant because the wind had been from the northwest for several hours prior to the incident. It is recommended that the reliability of data acquisition be improved. A study such as this one depends on valid, complete data it is of little value to collect data which has uncertainties as to time, calibration, or scale zeroes.