Present observations for assessing man's effect on the climate are inadequate. Although global temperature and radiosonde records are beginning to fill data needs, precipitation records in and near urban heat islands and major industrial zones with heat rejection and nuclei production are not sufficiently dense. Atmospheric observations of carbon dioxide (CO₂) are satisfactory, but more information on CO₂ sinks in the ocean and the biomass are needed. Atmospheric levels of other infrared absorbers have to be systematically watched. The total particulate load is a useless measure and needs to be replaced by regular and widespread observations of the size spectra of particulate matter. Adequate inventories of anthropogenic heat and water vapor emissions are required. Better inventories of man-made cloudiness have to be established. Additional atmospheric electric observations are also needed.