Urban effects on climate result from alteration of the energy, mass and momentum transfers of the pre-existing landscape by urban development. Results are presented from simultaneous suburban and nearby rural surface energy balance measurements, for the cities of Tucson, AZ, Sacramento, CA and Vancouver, B.C., using the rural measurements as a surrogate reference for the pre-urban conditions. The three cities have very different rural surroundings: desert (Tucson), semi-arid and irrigated (Sacramento), and moist farmland (Vancouver) due to their natural or managed land cover. These produce very different heat and water balance regimes. On the other hand, the energetic partitioning for the three suburban areas is similar, at least in relative terms. Therefore, urban-rural differences of energy and water transfers and climatic variables, often interpreted to be gauges of urban effects on climate, are really only "apparent". This follows because urban effects assessed in this manner are largely controlled by the state of the surrounding rural area - especially its relative wetness or dryness - rather than the nature of the urban area. These controls show up in the magnitudes of the related urban heat island and urban-rural humidity differences and provide another caution to those using urban-rural station pairs to assess possible contamination of long-term records by human activities.