Objectives: To compare the effect of treat-to-target-based escalations in conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) and biologics on clinical disease activity and magnetic resonance imaging (MRI) inflammation in a rheumatoid arthritis (RA) cohort in clinical remission. Method: One-hundred patients with established RA, Disease Activity Score based on 28-joint count–C-reactive protein (DAS28-CRP) < 3.2, and no swollen joints (hereafter referred to as ‘in clinical remission’) who received csDMARDs underwent clinical evaluation and MRI of the wrist and second to fifth metacarpophalangeal joints every 4 months. They followed a 2 year MRI treatment strategy targeting DAS28-CRP ≤ 3.2, no swollen joints, and absence of MRI osteitis, with predefined algorithmic treatment escalation: first: increase in csDMARDs; second: adding a biologic; third: switch biologic. MRI osteitis and Health Assessment Questionnaire (HAQ) (co-primary outcomes) and MRI combined inflammation and Simplified Disease Activity Index (SDAI) (key secondary outcomes) were assessed 4 months after treatment change and expressed as estimates of group differences. Statistical analyses were based on the intention-to-treat population analysed using repeated-measures mixed models. Escalation to first biologic compared to csDMARD escalation more effectively reduced MRI osteitis (difference between least squares means 1.8, 95% confidence interval 1.0–2.6), HAQ score (0.08, 0.03–0.1), MRI combined inflammation (2.5, 0.9–4.1), and SDAI scores (2.7, 1.9–3.5). Treat-to-target-based treatment escalations to biologics compared to escalation in csDMARDs more effectively improved MRI inflammation, physical function, and clinical disease activity in patients with established RA in clinical remission. Treatment escalation in RA patients in clinical remission reduces clinical and MRI-assessed disease activity. Trial registration: Clinicaltrials.gov identifier: NCT01656278.