The difficulties in defining the natural history of primary IgA nephropathy (IgAN) depend upon the pre-selection of patients for renal biopsy, a true individual variability - ranging from asymptomatic to rapidly progressive forms - as well as the use of different classifications of the renal lesions and statistical analyses sometimes carrying incorrect modalities. Long-term natural history studies have demonstrated that the rate of progression has an extremely wide range, from 5 to 25% after 10 years and 25-50% at 20 years, and complete remission is reported as well in 5 to 30% of cases. A geographic variability has been confirmed in a tri-continental study, explainable only partly by the earlier referral. Among the factors predicting progression, the more frequent in cohorts showing worse actuarial survival at 10 years are those associated with the advanced phases of renal damage, as increased creatinine level, arterial hypertension and nephrotic range proteinuria. A multivariate statistical approach showed the relevance of proteinuria during follow-up (percent duration of massive proteinuria or proteinuria at 1 year) more than proteinuria at the onset. Mean blood pressure value (MAP) and proteinuria during follow-up were independent predictors of end-stage CKD. Note the predictive value of severe microscopic hematuria in several studies. As far as histological features are concerned, strong independent predictors of progression at Cox multivariate analysis are the severity of glomerular sclerosis and interstitial fibrosis. The presence of crescents was a risk factor in almost all studies at univariate analysis, but did not maintain a significant predictor value at multivariate analysis. Conversely the association between crescents and tuft adhesions, possibly resulting from previous segmental necrosis, was found to be a significant risk factor. The extent of mesangial proliferation and parietal expansion of deposits was not significantly associated to unfavourable prognosis at multivariate analysis. The analysis of risk factors for progression of IgAN related to Henoch-Schoenlein purpura (HSP) failed to demonstrate any prognostic value for the presence and severity of extra-renal signs of vasculitis or presence of triggering factors. At multivariate Cox analysis, age and mean proteinuria during follow-up were powerful independent prognostic predictors. Proteinuria at baseline was not significantly related to renal progression, nor were hypertension or impaired renal function at onset. It is of interest that data at onset and at renal biopsy (renal function impairment, hypertension, nephrotic-range proteinuira) were not significantly related with renal detrimental progression. Neither had prognostic value the finding of crescents involving up to 75% of glomeruli.