The contribution of mitochondria to muscle adaptation to exercise is well known to all physiologists working on muscle plasticity and training. Since the classical paper by Holloszy (1967), a number of studies have demonstrated that endurance training can induce increases in mitochondrial density and in respiratory capacity of muscle fibres. More recently, the progress of knowledge in the field has delineated the signal chain based on AMPK and on PGC-1α and demonstrated the specific role of PGC-1α as the master gene able to activate mitochondrial biogenesis (see the recent review by Hawley et al. 2014). However, increasing amounts of evidence support the view that mitochondria are also involved in muscle mass regulation. The finding that PGC-1α overexpression protects against atrophy induced by denervation, fasting and other interventions through inhibition of FoxO gene expression (Sandri et al. 2006) was further corroborated by the observation that inhibition of mitochondrial fission has a similar protective or anti-atrophic effect (Romanello et al. 2010). This is in accordance with the view that the family of genes controlled by PGC-1α includes not only genes relevant for mitochondrial biogenesis but also genes which shift the balance between fission and fusion towards enhanced mitochondrial fusion. In muscle atrophy induced by denervation and fasting, the increased mitochondrial fission leads to a state of energy stress which activates AMPK and this in turn switches on FoxO3 and FoxO3-dependent transcription of atrogenes (Romanello et al. 2010). The question of whether such a mechanism could play a part in determining atrophy induced by disuse remains unanswered. In the present issue of The Journal of Physiology a paper by Cannavino and coworkers (2015) further strengthens the connection between mitochondrial impairment and muscle atrophy and gives an answer to the question concerning muscle atrophy in response to disuse. The paper represents the continuation of a previous study published by the same group (Cannavino et al. 2015). The model adopted in both studies was the ‘hindlimb suspension’ model in the mouse. The removal of load and the reduction of contractile activity can produce in a short time a very pronounced atrophy, in the order of 15% in less than one week. Starting from the widely accepted view that oxidative stress is a major cause of muscle degradation in disuse atrophy, the authors explored the response to an antioxidant treatment (administration of Trolox). Trolox was sufficient to control ROS accumulation and the specific antioxidant enzyme expression, but could not avoid the loss of muscle mass. In skeletal muscle fibres, ROS can derive from mitochondria as well as from other sources. The possible origin from mitochondrial metabolic impairment prompted the authors to analyse the mitochondrial metabolic function and mitochondrial dynamics. This analysis revealed that the mitochondria metabolic function was impaired and the balance between fission and fusion shifted towards fission. The demonstration of a causative connection between mitochondria and atrophy development was given by PGC-1α overexpression which at the same time prevented atrophy and restored the balance of the mitochondrial dynamics. It is worth emphasising the diversity among fast and slow muscles (see Fig. 1). In the slow soleus the primary event was the downregulation of PGC-1α followed by atrogene induction through FoxO3 disinhibition (Cannavino et al. 2015). In contrast, in the fast gastrocnemius the primary event was altered mitochondrial dynamics caused by downregulation of pro-fusion proteins and followed by energy stress and atrogene induction through AMPK activation. AMPK activation was sufficient to preserve PGC-1α expression (Cannavino et al. 2014). Importantly, in slow and fast muscles, the loss of mitochondria not only implies an impairment of aerobic metabolism, but also represents an ominous sign of imminent mass loss. Figure 1 Schematic representation of the different mechanisms leading to muscle atrophy in slow (soleus) and fast (gastrocnemius) muscles in the early phase of hindlimb suspension The results obtained by Cannavino and coworkers (2015 and 2014) are an important advance in our understanding of muscle adaptation to exercise and disuse and, additionally, represent a starting point for future work. New questions await an answer, such as, for example, the link between reduced loading and PGC-1α down-regulation or between reduced loading and mitochondria fission, the complex regulatory role played by AMPK and the definition of the retrograde signals that come from mitochondria and affect nuclear gene expression.