Non-invasive brain stimulation techniques (NIBS), such as Theta Burst Stimulation (TBS), Paired Associative Stimulation (PAS) and transcranial Direct Current Stimulation (tDCS), are widely used to probe plasticity in the human motor cortex (M1). Although TBS, PAS and tDCS differ in terms of physiological mechanisms responsible for experimentally-induced cortical plasticity, they all share the ability to elicit long-term potentiation (LTP) and depression (LTD) in M1. However, NIBS techniques are all affected by relevant variability in intra- and inter-subject responses. A growing number of factors contributing to NIBS variability have been recently identified and reported. In this review, we have readdressed the issue of variability in human NIBS studies. We have first briefly discussed the physiological mechanisms responsible for TBS, PAS and tDCS-induced cortical plasticity. Then, we have provided statistical measures of intra- and inter-subject variability, as calculated in previous studies. Finally, we have reported in detail known sources of variability by categorizing them into physiological, technical and statistical factors. Improving knowledge about sources of variability could lead to relevant advances in designing new tailored NIBS protocols in physiological and pathological conditions.