The current definition of rapid eye movement (REM) sleep without atonia has no quantitative character, and cut-off values above which the level of electromyographic tone can be considered to be ‘excessive’ are unclear. The aim of this study was to analyse the characteristics of chin electromyographic amplitude by means of an automatic approach in a large group of normal controls, subdivided into different age groups. Eighty-eight normal controls were included, subdivided into six age groups: preschoolers (≤6 years); schoolers (6–10 years); preadolescents (10–13 years); young adults (24–40 years); middle-aged (58–65 years); and old (>65 years). The average amplitude of the rectified submentalis muscle electromyographic signal was used for the computation of the REM sleep Atonia Index. Chin muscle activations were detected, and their amplitude, duration and interval analysed. REM sleep Atonia Index showed a progressive and rapid increase from the preschool age to school and preadolescent age, reaching the maximum in the young adult group; after this age a small decline was observed in the middle-aged and old subjects. Conversely, the number of movements per hour in REM sleep showed a ‘U’-shaped distribution across these age groups, with the minimum in the preadolescent group and the two extremes (preschool age and old) showing similar average levels of activity. Our results show that REM sleep atonia develops continuously during the lifespan, and undergoes complex changes with different developmental trajectories for REM atonia and electromyographic activations during REM sleep. Different mechanisms might subserve these two phenomena and their differential developmental dynamics.