OBJECTIVE To assess the mortality rate (MR) and the mortality risk of a rheumatoid arthritis (RA) inception cohort, with and without biologic agents (BAs). Other factors associated to mortality were also investigated. METHODS Retrospective longitudinal study of RA patients, attending the rheumatology outpatient clinic of a tertiary Hospital (Madrid), collected over 5 years (2000-2004), and followed from the diagnosis of RA up to the patients' death, lost to follow-up or September 2013. The dependent variable was death and the independent variable was exposure to BAs. Covariables: sociodemographic, clinical and therapy variables. MR was expressed per 1,000 patient-years with the 95% confidence interval [CI]. BA influence on MR was analysed by multivariable Cox models. Clinical and therapy variables were used in a time-dependent manner. The results are expressed in hazard ratio (HR) and [CI]. RESULTS We included 576 patients and 711 courses of therapy. 19.6% were taking BA, 86% disease-modifying anti-rheumatic drugs (DMARDs) (70% on methotrexate - MTX), and 12% were untreated. There were 133 deaths during 4,981.64 patient-years at risk. The MR for BA was 12.6 [6-26], for DMARDs was 22.3 [18.4-27.1], and for those without treatment was 89.1 [61.9-128.2]. The adjusted HR for mortality in those exposed to BA versus those not exposed was 0.75 [0.32-1.71]). Other variables independently associated with mortality were: age, rheumatoid factor, hospital admissions, Health Assessment Questionnaire (HAQ), and MTX use (HR: 0.44 [0.29-0.66]). CONCLUSIONS BAs and standard DMARDs are more effective in decreasing mortality compared to no therapy. Patients exposed to Bas were not associated with a significant increase or decrease in mortality when compared to patients with non-biological DMARDs. The use of MTX remains the only drug that has independently shown a beneficial effect on mortality.