Spatiotemporal channel modeling is fundamental to the design and implementation of future high-speed wireless communication systems. This paper investigates a two-dimensional--three-dimensional (2-D/3-D) hybrid ray-tracing method for outdoor microcellular and picocellular urban environments. It also explores the spatiotemporal channel model derived from this ray-tracing method. This 2-D/3-D ray-tracing method launches rays in a horizontal plane and establishes ray paths in three-dimensional (3-D) space. Partition vector incorporates other acceleration techniques to produce a method that reduces tracing time. The conventional verification of path loss and delay profiles predicted by ray tracing was extended to include the verification of angle of arrival (AOA). The results prove the necessity of extension and verify that path loss and delay profiles cannot fully confirm the accuracy of ray tracing in a spatial domain. A good level of accuracy was observed in the comparison of three key areas: (1) path loss, (2) delay profile, and (3) AOA. Degradation of prediction accuracy caused by incorrectly predicted phases in ray tracing was investigated, and improvements were made by incorporating a random-phase approach. It was found that measured path loss and delay profiles are almost fully confined within the 90% confidence interval, proving that the approach can account for the degradation effects caused by inaccurate antenna positions. The applicability of limited-band and limited beam width channel parameters transformed from ray-tracing results is also examined in the sense of an azimuth delay profile. The confirmation of the accuracy of the profile with measurement prepares ground for the use of ray-tracing approaches to analyze system performance in real environments. ¿ 2001 American Geophysical Union