Volunteer-based audio surveys are used to document long-term trends in anuran community composition and abundance. Current sampling protocols, however, are not region- or species-specific and may not detect relatively rare or audibly cryptic species. We used automated audio recording devices to record calling anurans during 2006–2009 at wetlands in Maine, USA. We identified species calling, chorus intensity, time of day, and environmental variables when each species was calling and developed logistic and generalized mixed models to determine the time interval and environmental variables that optimize detection of each species during peak calling periods. We detected eight of nine anurans documented in Maine. Individual recordings selected from the sampling period (0.5 h past sunset to 0100 h) described in the North American Amphibian Monitoring Program (NAAMP) detected fewer species than were detected in recordings from 30 min past sunset until sunrise. Time of maximum detection of presence and full chorusing for three species (green frogs, mink frogs, pickerel frogs) occurred after the NAAMP sampling end time (0100 h). The NAAMP protocol’s sampling period may result in omissions and misclassifications of chorus sizes for certain species. These potential errors should be considered when interpreting trends generated from standardized anuran audio surveys.