Biological drugs targeting pro-inflammatory or co-stimulatory molecules or depleting lymphocyte subsets made a revolution in rheumatoid arthritis (RA) treatment. Their comparable efficacy in clinical trials raised the point of the heterogeneity of RA pathogenesis, suggesting that we are dealing with a syndrome rather than with a single disease. Several tumor necrosis factor-alpha (TNF-α) blockers are available, and a burning question is whether they are biosimilar or not. The evidence of diverse biological effects in vitro is in line with the fact that a lack of efficacy to one TNF-α agent does not imply a non-response to another one. As proteins, biologicals are potentially immunogenic. It has been recently raised that anti-drug antibodies (ADA) may affect their bioavailability and eventually the clinical efficacy through local formation of immune complexes and directly by preventing the interaction between the drug and TNF-α. Regular monitoring of drug and ADA levels appears the best way to tailor anti-TNF-α therapies. Owing to the pleiotropic characteristics of the target, anti-TNF-α blockers may affect several mechanisms beyond rheumatoid synovitis. As TNF-α plays a pivotal role in the induction of early atherosclerosis, treatment with TNF-inhibitors may modulate cholesterol handling, in particular, cholesterol efflux from macrophages. Side effects are a major issue because of the systemic TNF-α blocking action. The efficacy of an anti-C5 monoclonal antibody fused to a peptide targeting inflamed synovia in experimental arthritis opened the way for new strategies: Homing to the synovium of molecules neutralizing TNF would allow to maximize the therapeutic action avoiding the side effects.