Purpose: To investigate whether exposure to pulse modulated radiofrequency (PM RF) influences human cognitive performance, and whether it does so in a dose-dependent manner. Materials and methods: Thirty-six healthy adults participated in a randomized, double-blind, counterbalanced provocation study. Cognitive performance was assessed using a visual discrimination task and a modified Sternberg working memory task, which were calibrated to individual performance levels in a preliminary testing session. An sXh920 planar exposure system was used to generate a 920 MHz GSM-like signal, providing three conditions (peak-spatial SAR averaged over 10 g) of 0 W/kg (sham), 1 W/kg (low RF) and 2 W/kg (high RF). Results: A significant decrease in reaction time (RT) in the Sternberg working memory task was found during exposure compared to sham. This effect was not dose-dependent. Conclusions: Cognitive performance was shown to be faster under PM RF conditions, relative to sham, in a working memory task. While the majority of the literature has not found effects of PM RF exposure on cognitive performance, it is possible that the methodological improvements employed in the present study increased sensitivity, and thus the ability to detect potential effects.