Purpose: To evaluate the influence of substrate moisture on the clinical behavior of two dentin adhesives after 18 months. The null hypothesis tested was that drying dentin with air upon rinsing off the acid would not influence the clinical performance of two dentin adhesives as compared to leaving the preparation visibly moist. Materials and Methods: Thirty-five patients were enrolled in this study. One hundred twenty-eight restorations divided into 4 groups were inserted and evaluated at baseline: (1) NT/Moist - Prime & Bond NT, an acetone-based adhesive, applied on moist dentin; (2) NT/Dried - Prime & Bond NT applied on dentin dried with air for 3 to 4 s; (3) SB/Moist - Single Bond, an ethanol- and water-based adhesive, applied on moist dentin; (4) SB/Dried - Single Bond applied on dentin dried with air for 3 to 4 s. A microfilled composite resin was used for all restorations. Patients were recalled at 6 and 18 months. Results: At 18 months after initial placement, 110 restorations (86% recall rate) were re-evaluated. Retention rates at 18 months were 92% for NT/Moist, 93% for NT/Dried, 100% for SB/Moist, and 89% for SB/Dried. No statistically significant differences were found among groups for retention rate. Both NT/Moist and SB/Moist resulted in a significant decrease in sensitivity to air from baseline to 18 months. When data were pooled for the variable "substrate moisture", SB resulted in an overall retention rate of 95%, while NT resulted in a retention rate of 92% (statistically similar). The marginal adaptation with SB was significantly worse at 18 months than at baseline. Conclusion: The moisture level of the dentin substrate in noncarious cervical lesions does not influence retention of composite restorations, but moist conditions caused less sensitivity to air. When applied as per manufacturers' instructions (moist dentin), both adhesives resulted in Class V retention rates exceeding the ADA 18-month full acceptance guidelines. Keywords: dental bonding, clinical trial, composite restorations