Intra-articular (IA) corticosteroids relieve osteoarthritis (OA) pain, but rapid absorption into systemic circulation may limit efficacy and produce untoward effects. We compared the pharmacokinetics (PK) of IA triamcinolone acetonide (TA) delivered as an extended-release, microsphere-based formulation (FX006) vs a crystalline suspension (TAcs) in knee OA patients. This Phase 2 open-label study sequentially enrolled 81 patients who received a single IA injection of FX006 (5 mL, 32 mg delivered dose, N = 63) or TAcs (1 mL, 40 mg, N = 18). Synovial fluid (SF) aspiration was attempted in each patient at baseline and one post-IA-injection visit (FX006: Week 1, Week 6, Week 12, Week 16 or Week 20; TAcs: Week 6). Blood was collected at baseline and multiple post-injection times. TA concentrations (validated LC-MS/MS, geometric means (GMs)), PK (non-compartmental analysis models), and adverse events (AEs) were assessed. SF TA concentrations following FX006 were quantifiable through Week 12 (pg/mL: 231,328.9 at Week 1; 3590.0 at Week 6; 290.6 at Week 12); post-TAcs, only two of eight patients had quantifiable SF TA at Week 6 (7.7 pg/mL). Following FX006, plasma TA gradually increased to peak (836.4 pg/mL) over 24 h and slowly declined to <110 pg/mL over Weeks 12-20; following TAcs, plasma TA peaked at 4 h (9628.8 pg/mL), decreased to 4991.1 pg/mL at 24 h, and was 149.4 pg/mL at Week 6, the last post-treatment time point assessed. AEs were similar between groups. In knee OA patients, microsphere-based TA delivery via a single IA injection prolonged SF joint residency, diminished peak plasma levels, and thus reduced systemic TA exposure relative to TAcs.