Universal mobile telecommunication system (UMTS) networks should be deployed according to cost-effective strategies that optimize a cost objective and satisfy target quality-of-service (QoS) requirements. In this paper, we propose novel algorithms for joint uplink/downlink UMTS radio planning with the objective of minimizing total power consumption in the network. Specifically, we define two components of the radio planning problem: 1) continuous-based site placement and 2) integer-based site selection. In the site-placement problem, our goal is to find the optimal locations of UMTS base stations (BSs) in a certain geographic area with a given user distribution to minimize the total power expenditure such that a satisfactory level of downlink and uplink signal-to-interference ratio (SIR) is maintained with bounded outage constraints. We model the problem as a constrained optimization problem with SIR-based uplink and downlink power control scheme. An algorithm is proposed and implemented using pattern search techniques for derivative-free optimization with augmented Lagrange multiplier estimates to support general constraints. In the site-selection problem, we aim to select the minimum set of BSs from a fixed set of candidate sites that satisfies quality and outage constraints. We develop an efficient elimination algorithm by proposing a method for classifying BSs that are critical for network coverage and QoS. Finally, the problem is reformulated to take care of location constraints whereby the placement of BSs in a subset of the deployment area is not permitted due to, e.g., private property limitations or electromagnetic radiation constraints. Experimental results and optimal tradeoff curves are presented and analyzed for various scenarios.