Opportunistic selection improves the performance of a multi-node wireless system by exploiting multi-user or spatial diversity. In it, the nodes are sorted in the descending order of their metrics, which captures the utility of a node to the system if selected, and the best node with the highest metric is selected. We analyze the effect of imperfect power control on the conventional timer with power control scheme, which selects the best node in a distributed manner, and quantify the extent by which it reduces the probability of selecting the best node and increases the probability of selecting a non-best node. We then redesign it to ameliorate the impact of imperfect power control. Our systematic approach eschews several ad hoc assumptions implicit in the design of the conventional timer scheme, and jointly optimizes its various parameters to maximize the probability of selecting the best node in the presence of imperfect power control. We present several structural insights, including asymptotic ones, about the optimal scheme, which also enable it to be determined with much lower computational complexity. Our benchmarking results show that it is scalable and outperforms the conventional schemes.