Observed changes in the frequency of intense precipitation and of wet days have been analyzed for more than half of the land area of the globe. These observed changes have been analyzed in relation to changes in intense precipitation for three transient Global Climate Model (GCM) simulations, each with greenhouse gas concentrations increasing during the 20th and 21st centuries and doubling during the later part of the 21st Century. We found that the empirical evidence from the period of instrumental observations and the model projections of a greenhouse-enriched atmosphere both indicate an increasing probability of intense precipitation events for many extra-tropical regions including the United States. In several regions of the world, rainy days are becoming less frequent. These two patterns may occur in the same regions where an increasing probability of intense precipitation with a decreasing frequency of rainy days makes changes in mean precipitation less notable then changes in precipitation distribution spectra and also affecting the entire terrestrial hydrological cycle (runoff, soil moisture available for evapotranspiration, etc.). In the northern regions, observed increases in surface air temperatures have resulted in (a) an earlier snow cover retreat, (b) an increase in the frequency of cold season thaws, and (c) a significant expansion of the warm season. This coupled with changes in precipitation distribution spectra, have resulted in the possibility for a “summer dryness” scenario in several large interior regions of the continents (cf., recent forest fires and droughts in high latitudes such as Siberia, Central Asia, and several regions of the western part of North America). These conditions were projected by some GCMs for the greenhouse gases enriched atmosphere and suggest reduced future water availability. Reliable assertions of trends in some of characteristics of extremes (e.g., very heavy and extreme precipitation changes) are possible only for those regions with dense networks due to the restricted radius of correlation for these characteristics. Therefore, an assessment of the representativeness of results based on sparse networks has been a part of the analyses and will be presented at the conference.