The history of deep brain stimulation for Parkinson's disease (PD) represented a paradigmatic cross-talk between mammalian disease models and clinical evidence in humans. Fascinating were the results achieved by high frequency stimulation (HFS) into the subthalamic nucleus (STN) of MPTP-treated primates. An analogous strategy relieved tremor and hypokinetic parameters in PD patients. The 6-hydroxydopamine (6-OHDA) rodent model has mastered decades of research, contributing to understanding of the PD pathology. However, this review wonders about the actual synergy between the routine neurotoxic models and PD patients underlying STN-DBS. At first, some findings collected following 6-OHDA, promoted dogmatic visions, as the wrong contention that suppression of STN glutamate was the key therapeutic player. Instead, changes of glutamate release are negligible in humans during transition to ON-state. Besides, the imbalance of basal ganglia endogenous band frequencies, the beta (β) band increase and the cortical-basal ganglia synchronization, undisputedly shared by models and PD patients, do not govern the whole spectrum of non-motor PD signs, difficult to investigate in rodents. Furthermore, the tonic release of dopamine, inferred during HFS in rodents, was not replicated in humans. Finally, neurotoxic rodent models describe a 'pure' dopamine depletion sparing pathways crucial in parkinsonian phenotypes, that is, noradrenergic and cholinergic ones. Although the utilization of neurotoxic models is still providing major advancements, we pore over these contradictions and try to support possible amendments of neurotoxic models (advocating modern 'in vivo' approaches and recordings extending towards motor thalamus) for pursing the development of new DBS technology.