Purpose: The purpose of this study was to evaluate the role of keratinised mucosa on the long-term success of dental implants. Materials and methods: Thirty-two edentulous patients had one of their dental arches rehabilitated with a provisional screw-retained resin reinforced cross-arch fixed prosthesis supported by four immediately loaded implants. The two central straight implants were randomly allocated in two equal groups, according to a parallel-group design, to receive or not intermediate abutments (Multi-Unit Abutment, MUA). However, for the purpose of the present publication, the study was considered as a prospective single cohort study. To be immediately loaded, implants had to be inserted with a minimum torque of 30 Ncm, which was achieved by all implants. Provisional prostheses were delivered within 24 hours and were replaced, after 4 months, by definitive screw-retained metal-ceramic prostheses. Patients were followed up to 5 years after loading. Prostheses were removed every 8 months to facilitate professionally delivered maintenance. Outcome measures were prosthesis and implant failures, complications, peri-implant marginal bone level (MBL) changes, bleeding on probing (BoP) and keratinised mucosa height (KMH). The absence of vestibular or lingual keratinised mucosa was put in relation with peri-implant bone loss and BoP. Results: Five-years after loading two patients dropped out and no implant or prosthetic failure occurred. Four patients were affected by prosthetic complications. At delivery of definitive prostheses, 32 (25%) implants had at least one vestibular or lingual site with no KMH and 96 (75%) implants had both sites with some KMH. Multi-level models did not show any statistically significant association between KMH at the time of delivery of the definitive prosthesis and changes in MBL and BoP at 5 years (estimate = 0.02; 95% CI: -0.02 to 0.05; P = 0.3393; and 1.02; 95% CI: 0.79 to 1.33; P = 0.8695, respectively). When KMH was analysed as dichotomous variable, implants with presence of KMH at delivery of definitive prosthesis at both vestibular and lingual aspects showed a trend of less BoP (estimate = -0.8; 95% CI: -1.69 to 0.08; P = 0.0741) but a statistically significant greater MBL loss compared to implants where KMH was only present at one site (estimate = 0.18; 95% CI: -0.1 to 0.3; P = 0.0041). Conclusions: The 5-year after loading outcome of immediately loaded screw-retained cross-arch prostheses supported by four implants is excellent in both the maxilla and mandible. Although the height of the keratinised mucosa did not seem to alter the clinical outcomes, its presence both at vestibular and lingual sites was associated with an increased marginal bone loss when compared to implants having at least one side without keratinised mucosa. Keywords: dental implants, full edentulism, immediate loading, keratinised mucosa Conflict of interest statement: This study was completely self-funded and the authors declare not to have any conflict of interest.