Physical impairments, such as noise and signal distortions, negatively affect the quality of information transfer in optical networks. The effect of physical impairments predominantly augments with distance and bit rate of the signal to the point that it becomes detrimental to the information transfer. To reverse the effect of physical impairments, the signal needs to be regenerated at nodes that have regeneration capabilities. Regenerators are costly and are, therefore, usually only sparsely placed in the network, in which case it is referred to as a translucent network. This paper deals with two problems in translucent networks, namely: (1) how to incorporate impairment awareness in the routing algorithms, and (2) how many regenerators to place inside the network and where. We propose exact and heuristic algorithms for impairment-aware path selection and, through simulations, show that our heuristic T IARA is computationally efficient and performs very close to our exact algorithm EIARA. Subsequently, we propose a greedy algorithm for placing regenerators that, contrary to previous proposals, is suitable for multiple impairment metrics, has polynomial complexity for a single impairment metric, and is cheaper in terms of the number of regenerators needed.