To further investigate into the epidaemiology of systemic lupus erythematosus (SLE) in Southern Europe, we have assessed the incidence, clinical spectrum and survival of patients diagnosed with late-onset SLE (age ≥ 50 years) according to the 1982 American College of Rheumatology (ACR) classification criteria at the single hospital for a well-defined population of Lugo, Northwestern (NW) Spain. Between January 1987 and December 2006, 51 (39.3%) of the 150 patients diagnosed as having SLE fulfilled definitions for late-onset SLE. The predominance of women among late-onset SLE (4:1) was reduced when compared with that observed in early-onset SLE (7:1). However, the incidence of late-onset SLE was significantly higher in women (4.2 [95% confidence interval (CI): 3.1-5.6] per 100,000 population) than in men (1.3 [95% CI: 0.6-2.2] per 100,000 population) (p < 0.001). As observed in early-onset SLE, the most frequent clinical manifestation in patients with late-onset SLE was arthritis (71.2%). Renal disease was less common in late-onset SLE (13.5%) than in early-onset SLE (26.4%); p = 0.07). In contrast, secondary Sjögren syndrome was more commonly found in the older age-group (27.1% versus 12.1%; p = 0.03). A non-significantly increased incidence of serositis was also observed in late-onset SLE patients (33.9% versus 22.0%; p = 0.13). Hypocomplementaemia (72.9% versus 91.2%) and positive results for anti-DNA and anti-Sm (49.2% and 6.8% versus 68.1% and 23.1, respectively) were significantly less common in late-onset SLE patients than in early-onset SLE. The probability of survival was reduced in late-onset SLE (p < 0.001). With respect to this, the 10-year and 15-year survival probability were 74.9 % and 63.3% in the late-onset SLE group and 96.3% and 91.0% in patients with early-onset SLE, respectively. In conclusion, our results confirm that in NW Spain SLE is not uncommon in individuals 50 years and older. In keeping with earlier studies, late-onset SLE patients from NW Spain have some clinical and laboratory differences with respect to those individuals with early-onset SLE. Our data support the claim of a reduced probability of survival in the older age-group of SLE patients.