Purpose: To compare the long-term survival rates of smooth- and rough-surface dental implants among smokers and nonsmokers. Materials and Methods: A retrospective chart review was conducted for 2 time periods: January 1, 1991, through December 31, 1996, during which smooth-surface implants were utilized, and January 1, 2001, through December 31, 2005, during which roughsurface implants were utilized. This review included all implants placed and restored in 1 institution during the 2 timeframes. Data were specifically collected relative to patient age, gender, smoking status, implant diameter, implant length, and anatomic location of implants. Implants from the first and second time periods were followed through mid-1998 and mid-2007, respectively. Associations of patient/implant characteristics with implant survival were evaluated using marginal Cox proportional hazards models (adjusted for age and gender) and summarized with hazard ratios (HR) and corresponding 95% confidence intervals (CI). Results: A total of 593 patients (322 [54.3%] female; mean [SD] age, 51.3 [18.5] years) received 2,182 smooth-surface implants between 1991 and 1996, while 905 patients (539 [59.6%)] female; mean [SD] age, 48.2 [17.8] years) received 2,425 rough-surface implants between 2001 and 2005. Among the rough-surface implants, smoking was not identified as significantly associated with implant failure (HR = 0.8; 95% CI = 0.3 to 2.1; P = .68). In contrast, smoking was associated with implant failure among the group with smooth-surface implants (HR = 3.1; 95% CI = 1.6 to 5.9; P .001). Implant anatomic location was not associated with implant survival among patients with rough-surface implants (P = .45) and among nonsmokers with smooth-surface implants (P = .17). However, anatomic location affected the implant survival among smokers with smooth-surface implants (P = .004). In particular, implant survival was the poorest for implants placed in the maxillary posterior areas of smokers. Conclusions: Based on this retrospective study, the following observations were made: Smoking was identified as a risk factor for implant failure of smooth-surface implants only; among the smokers who received smooth-surface implants, an association was identified between implant failure and location of the implant placement; no association was identified between implant failure and location among the smokers who received rough-surface implants. Palabras clave: dental implants, rough surface implants, smooth surface implants, survival rates