Relationships between viruses and autoimmune diseases such as systemic lupus erythematosus (SLE) are still elusive. Recent reports demonstrated the association of some viral infections with peculiar clinical events in the general population, such as cytomegalovirus (CMV) with arterial damage and Parvovirus B19 (PV-B19) with hematologic abnormalities. We planned to look for this kind of viral imprinting in SLE, hypothesizing that traces of specific features of some viral infections might be found in some subsets of seropositive SLE patients. In 60 SLE patients recruited at our nephrologic center, serology for CMV, PV-B19, Epstein-Barr virus viral capsid antigen (EBV-VCA), Epstein-Barr nuclear antigen (EBNA) and Epstein-Barr virus early antigen (EBV-EA) was performed. chi2 and ANOVA were employed to compare the frequency and titers of antiviral antibodies in SLE patients with groups of transplant, hemodialysis and blood donor subjects. chi2, Fisher's test, Bonferroni and Scheffe's test were employed to compare the different biochemical/clinical features between seropositive and seronegative SLE patients. Univariate and multivariate analysis (logistic regression models) were employed to evaluate the odds ratio (OR) of different risk factors for vascular events (including Raynaud's phenomenon, deep venous thrombosis) and hematologic abnormalities (including severe anemia, leukopenia and thrombocytopenia). Anti-CMV (82%), anti-PV-B19 (60%), anti-EBV-VCA (92%) and EBV-EA (45%) IgG antibodies were frequent in SLE, with higher prevalence in comparison with the blood donor group and higher titers in comparison with transplant and hemodialysis groups. CMV seropositivity was a highly significant risk factor for Raynaud's phenomenon (OR +alpha in univariate and multivariate analysis = 13.51 using a correction of 0.5 in case of a zero event), but not for venous vascular events (OR = 1.31). An increased though not significant risk factor was found for antiphospholipid antibodies (OR = 2.71, p = 0.19), while the presence of nephrotic syndrome during the follow-up was a significant protective factor (OR = 0.15, p = 0.035). There was no significantly increased OR for PV-B19 seropositivity in cases with severe anemia (OR = 2.09, p = 0. 29). No significant associations were found with the status of EBV reactivation. In conclusion, our results support the hypothesis that viral infection may imprint the course of SLE leading to specific clinical subsets (i.e. CMV and 'vascular' SLE, with more frequent Raynaud's phenomenon and a less frequent typical histological renal picture responsible for nephrotic syndrome). Further prospective studies are justified to validate these correlations, mainly dealing with associations between acute viral infections and vascular events, thus eventually leading to a better understanding of mutual relationships between viruses and SLE.