The importance of patent foramen ovale (PFO) in stroke of unknown cause remains disputed, as PFO may be present in up to 20% of normal people and in a high proportion of patients with non-vascular disorders. Recent evidence suggests that the amount of right-to-left shunt (RLS) may be the crucial factor for stroke occurrence and relapse. The aim of the study was to assess predictors of recurrence in PFO-related stroke patients with particular emphasis on amount of shunting. Patients less than 61 years old who had been admitted for a PFO-related stroke within the previous 5 years, were re-evaluated on a follow-up visit. The clinical syndrome, residual disability, vascular risk factors and number of relapses as the index event were assessed. RLS sizing was semi-quantitatively performed with saline-enhanced transcranial Doppler (TCD), by assuming a cut-off of more or less 10 bubbles recorded in the cerebral vessels as a criterion to discriminate large versus small shunt, respectively. Thereafter patients were prospectively followed-up for a median time of 23 months. Total follow-up was 61 months. Fifty-nine patients (M/F = 23/36, mean age 43 +/- 13) were studied. Overall there were 23 relapses in 13 patients. The amount of shunting was the only significant independent variable associated with relapse: at the end of the follow-up period the recurrence rate was 0.66 and 8.2% per patient per year in patients with small and large shunt, respectively. This difference was statistically significant (chi2 = 10.39, P = 0.0012; OR 17.05, 95% CI 2.10-755.22). In patients with PFO-related stroke, the amount of RLS as assessed with TCD is the only independent predictor of relapse. PFO sizing is mandatory in patients with PFO.