BACKGROUND/OBJECTIVE: Arthrocentesis of the noneffusive knee is challenging. We hypothesized (1) pneumatic compression of the noneffusive knee would improve arthrocentesis, and (2) the increased synovial fluid yield would be diagnostically useful. METHODS: One hundred and thirty consecutive patients with a painful but clinically noneffusive knee underwent arthrocentesis. Arthrocentesis was performed with and without pneumatic compression and fluid volumes from each were measured. Successful arthrocentesis was defined as a fluid yield ≥ 1.0 mL. Synovial fluid samples were analyzed for total nucleated cell (TNC) count, TNC differential, the presence of crystals by polarized microscopy, and bacteria by culture. RESULTS: Arthrocentesis of the noneffusive joint was successful ( ≥ 1.0 mL) in 15.4% (20/130) with conventional arthrocentesis and in 53.8% (70/130) with pneumatic compression (OR: 6.4; CI: 3.6-11.6; p=0.0001). Pneumatic compression increased synovial fluid yield: conventional arthrocentesis: 0.5±1.1 mL, compression: 2.8±3.5 mL [2.3±3.0 mL increase (+460%), 95% CI: -2.9<-2.3<-1.7 mL; p=0.0001]. Synovial fluids were abnormal in 60.0% (42/70) of aspirated knees. Mean TNC was 2746±10,741 cells/mm3, including 40.0% normal fluid (TNC/mm3<200), 49.2% noninflammatory fluid (TNC/mm3 ≥ 200 but <2000), 10.8% inflammatory fluid (TNC/mm3 ≥ 2000), 1.5% hemorrhagic fluid, and 10% fluids with crystals that changed the diagnosis and subsequent therapy. CONCLUSIONS: Pneumatic compression-assisted arthrocentesis enables the successful collection of synovial fluid in the clinically noneffusive symptomatic knee. The majority of extracted synovial fluids are abnormal, indicating inflammatory arthritis, noninflammatory arthritis, or previously unrecognized crystal-associated disease. Pneumatic compression-assisted arthrocentesis is useful in the evaluation of the clinically noneffusive symptomatic knee.