In the past 20 years of research regarding effects of mobile phone-derived electromagnetic fields (EMFs) on human cognition, attention has been one of the first and most extensively investigated functions. Different domains investigated covered selective, sustained, and divided attention. Here, the most relevant studies on this topic have been reviewed and discussed. A total of 43 studies are reported and summarized: of these, 31 indicated a total absence of statistically significant difference between real and sham signal, 9 showed a partial improvement of attentional performance (mainly increase in speed of performance and/or improvement of accuracy) as a function of real exposure, while the remaining 3 showed inconsistent results (i.e., increased speed in some tasks and slowing in others) or even a worsening in performance (reduced speed and/or deteriorated accuracy). These results are independent of the specific attentional domain investigated. This scenario allows to conclude that there is a substantial lack of evidence about a negative influence of non-ionizing radiations on attention functioning. Nonetheless, published literature is very heterogeneous under the point of view of methodology (type of signal, exposure time, blinding), dosimetry (accurate evaluation of specific absorption rate-SAR or emitted power), and statistical analyses, making arduous a conclusive generalization to everyday life. Some remarks and suggestions regarding future research are proposed.