Rationale: Studies have shown elevated inflammatory biomarkers in obstructive sleep apnea (OSA), but data after continuous positive airway pressure (CPAP) treatment are inconsistent. Objectives: We used the Olink proteomics panel to identify unique OSA clusters on the basis of inflammatory protein expression and assess the impact of CPAP therapy. Methods: Adults with newly diagnosed OSA had blood drawn at baseline and three to four months after CPAP. Samples were analyzed using the Olink proteomics platform, which measures 92 prespecified inflammatory proteins using proximity extension assay. Linear mixed-effects models were used to model changes in protein expression during the period of CPAP use, adjusting for batch, age, and sex. Unsupervised hierarchical clustering was performed to identify unique inflammatory OSA clusters on the basis of inflammatory biomarkers. Within-cluster impact of CPAP on inflammatory protein expression was assessed. Results: Among 46 patients, the mean age was 46 ± 12 years (22% women), mean body mass index was 31 ± 5 kg/m2, and mean respiratory disturbance index was 33 ± 17 events/hour. Unsupervised cluster and heatmap analysis revealed three unique proteomic clusters, with low (n = 21), intermediate (n = 19), and high (n = 6) inflammatory protein expression. After CPAP, there were significant within-cluster differences in protein expression. The low inflammatory cluster had a significant increase in protein expression (16%; P = 0.02), and the high inflammatory cluster had a significant decrease in protein expression (-20%; P = 0.003), more significant among those compliant with CPAP in the low (25%; P = 0.04) and high (-22%; P = 0.01) clusters. Conclusions: We identified three unique inflammatory clusters in patients with OSA using plasma proteomics, with a differential response to CPAP by cluster. Our results are hypothesis generating and require further investigation in larger longitudinal studies for enhanced cardiovascular risk profiling in OSA.