Audio recording systems coupled with automated song recognition are commonly being used for monitoring wildlife. Recorders usually differ in cost and effectiveness, and their performance may vary with source distance, wind speed and acoustic source direction, among other factors. We here assess the cost-effectiveness of five audio systems considering such factors as distance and singing direction. We developed field tests using playback of Dupont's Lark Chersophilus duponti songs from nine fixed locations at distances of 1 to 256m, played towards or away from the recorders' position. We selected this species because its very characteristic song should be easily identified by automated signal recognition software. Field tests were carried out during March 2016 in level dwarf-shrub steppe (mean height < 40cm) in NE Spain. We found large differences in effectiveness between recorders. The number of songs detected by an automated signal recognition algorithm significantly decreased with distance and when playback was angled away from the recorder position, a factor never previously tested. Finally, we give the design of a cost-effective Autonomous Recording Unit, based upon the most effective recorder. We recommend researchers working with acoustic recorders to evaluate the performance of several devices before making a selection for long-term monitoring programmes, and to consider such factors such as singing direction in their analyses. —pérez-Granados, C., Bota, G., Giralt, D., Albarracín, J. & Traba, J. (2019). Cost-effectiveness assessment of five audio recording systems for wildlife monitoring: differences between recording distances and singing direction. Ardeola, 66: 311-325.