Objective: To evaluate clinical, laboratory, or radiographic predictors of the onset of interstitial lung disease in systemic sclerosis. Methods: Sixty-five out of 220 systemic sclerosis outpatients, without interstitial lung disease at baseline and with ⩾3 chest high resolution computed tomography scans during follow-up were recruited. Thoracic lymphadenopathy and interstitial lung disease were assessed by chest high resolution computed tomography. Hazard ratio (95% confidence interval) of interstitial lung disease occurrence was assessed by Cox regression models, adjusting patient’s demographics and disease characteristics. Sensitivity, specificity, and accuracy of the interstitial lung disease predictors were evaluated by receiver operating characteristic analysis. Results: The development of interstitial lung disease was observed in 44/65 (68%) patients. Thoracic lymphadenopathies was detected in 40/65 (61%) patients, of whom 36 (82%) developed interstitial lung disease, but only four patients with thoracic lymphadenopathies did not develop ILD at last visit of follow-up (19%) (p = 0.0001). Adjusted hazard ratio of systemic sclerosis-interstitial lung disease onset was 5.8 (95% confidence interval, 2.0–16.5) for thoracic lymphadenopathy, which preceded by 108 ± 98 weeks the systemic sclerosis-interstitial lung disease detection. Thoracic lymphadenopathy had 84% specificity, 81% sensitivity, and 0.82 accuracy to predict interstitial lung disease. In particular, anticentromere antibodies or limited cutaneous subset of systemic sclerosis patients with thoracic lymphadenopathy showed earlier interstitial lung disease onset than those without lymphadenopathy. In addition, patients who developed interstitial lung disease had higher frequency of anti-Scl-70 (57% vs 19%; p = 0.009) and diffuse cutaneous subset (29% vs 3%; p = 0.02) than those who did not. Conclusions: Thoracic lymphadenopathy was the strongest independent predictor of systemic sclerosis-interstitial lung disease, mostly in anticentromere antibodies and limited cutaneous subset of systemic sclerosis patients. Further prospective studies are needed to confirm our preliminary data and to understand whether thoracic lymphadenopathies may have a pathogenetic role in interstitial lung disease development.