Sampling methods that can be implemented efficiently while minimizing bias are central for achieving strong statistical power in field studies. We estimated site occupancy and detection probabilities for Red-Backed Salamanders, Plethodon cinereus, using natural cover object (NCO) and leaf litter surveys and tested the power of these methods to detect occupancy differences in simulated data. Both NCO and leaf litter surveys performed well in their ability to detect P. cinereus (pˆ  =  0.59 ± 0.07 [± 1 SE]; 0.49 ± 0.08, respectively) and produced similar site occupancy estimates (ψ ˆ  =  0.75 ± 0.09; 0.77 ± 0.11, respectively); however, leaf litter searches took three times longer to complete. Our simulation study revealed the time required to complete leaf litter surveys limited the number of sampling sites, which reduced statistical power for detecting a 25% change in salamander occupancy between two groups of sites. Natural cover object searches provided greater power for detecting a similar change in occupancy, largely because of high sampling size. Although reliable salamander detection data can be collected quickly using NCO surveys, we discuss situations where this technique may not be appropriate. Our study highlights the importance of accounting for imperfect detection probability when conducting visual encounter surveys for terrestrial plethodontids.