Over the last decades, several computational (in silico) methods have been developed and applied to test pharmacological hypotheses. An important hypothesis is that the therapeutic activity of many commonly used antidepressants may be partially mediated through the inhibition of different nicotinic acetylcholine receptors (nAChRs). This is based on pathologic conditions where the activity of the cholinergic system is exacerbated compared to the adrenergic system. Different in silico methods, including comparative/homology modeling, molecular docking, and molecular dynamics simulations, have been employed to study the interactions between several classes of antidepressants with distinct nAChR subtypes. More specifically, these methods were used to structurally characterize the antidepressant binding sites and to better understand their inhibitory mechanisms. This review focuses on computational methods that were found important in explaining and supporting several experimental results concerning the interaction of antidepressants with different nAChR subtypes. Among the studied antidepressants are norepinephrine selective reuptake inhibitors [e.g., (-)-reboxetine] as well as less selective antidepressants such as dopamine/norepinephrine reuptake inhibitors [e.g., (±)-bupropion and its derivatives], tricyclic antidepressants (e.g., imipramine), and (±)-mecamylamine and its enantiomers.