This work presents theoretical predictions obtained by Blaunstein et al. <1997>, Blaunstein and Levin <1996>, Walfisch and Bertoni <1988>, Xia et al. <1993>, Xia and Bertoni <1992>, and Bertoni et al. <1994> for estimating microcell characteristics in a street environment in order to improve the effective service to individual subscribers to cellular wireless systems. In the case when the investigated region is built in a grid-like manner with regularly distributed rows of buildings and both antennas are located under the average building height in conditions of line of sight (LOS) along the street, we use the model of a three-dimensional (3-D) multislit waveguide, according to Blaunstein et al. <1997> and Blaunstein and Levin <1996>, to estimate the path loss at the street level. In the case of obstructive (NLOS) conditions, when both antennas are placed below and above the rooftops in an environment with strong shadowing surrounding them, we use the 2-D model of multidiffraction from the building roofs and walls, according to Walfisch and Bertoni <1988>, Xia et al. <1993>, Xia and Bertoni <1992>, and Bertoni et al. <1994>, modified in conjunction with actual variations of building heights, the distances between them, and the actual base station antenna height variations; at street intersections we use the crossing-waveguide model according to Blaunstein and Levin <1997>. The contributions to path loss are obtained to predict the experimentally observed coverage effects and a cell radius in LOS and NLOS conditions to predict the microcell shape for cellular map construction in urban and suburban areas with regularly distributed rows of buildings and rectangularly crossing streets. ¿ 1999 American Geophysical Union