The retention rate of a biological drug (percentage of patients remaining on treatment over time) provides an index of a drug's overall effectiveness. The golimumab retention rate as first-line biological therapy was high in clinical trial extensions lasting 5 years. Real-world studies also indicate good retention rates but have been of shorter duration. The probability of retention with golimumab treatment was assessed, as any line of anti-tumor necrosis factor-alpha therapy, for up to 5 years in patients with rheumatoid arthritis (RA), axial spondyloarthritis (SpA) or psoriatic arthritis (PsA), associated factors were analyzed. A retrospective database analysis of the Spanish registry of patients with rheumatic disorders receiving biological drugs (BIOBADASER) was performed. Among 353 patients, 29.8% had RA, 41.6% SpA and 28.6% PsA. Golimumab was the first biological drug in 40.1% of patients, second in 30.1% and third/later in 29.8%. The overall probability of retention of golimumab at years 1, 2, 3, 4 and 5 was 85.9% (95% confidence interval 81.4-89.5%), 73.7% (67.1-79.1%), 68.5% (60.5-75.1%), 60.6% (50.2-69.5%) and 57.1% (44.9-67.5%), respectively. Retention was similar across indications (p = 0.070) but was greater when golimumab was used as the first biological agent compared with later therapy lines (p < 0.001). Factors associated with higher retention of golimumab treatment (Cox regression) were use as a first-line biological and concomitant methotrexate treatment; corticosteroid need was associated with lower retention. The long-term probability of golimumab retention was high in this real-world study of patients with rheumatic diseases, especially when used as the first biological drug.