We consider the problem of designing a new natural gas transmission network or expanding an existing network while minimizing the total investment and operating costs. We develop an integrated large-scale mixed-integer nonlinear optimization model to determine pipelines in the network, compressor stations and their capacities, timings of these installations in a multi-period planning horizon, and natural gas purchase and steady-state flow decisions for each period in the network. The model is solvable using state-of-the-art solution methodologies available online. Employing our modeling and the solution methodology for its solution, we conduct computational studies on various test instances generated using realistic network structure and data based on the natural gas network in Turkey as well as data from literature. Our analysis provides insights into sensitivity of network configuration and operations to the number of periods within the planning horizon and cost parameters as well as into strategic decision making for design and expansion of natural gas networks. (C) 2014 Elsevier Ltd. All rights reserved.