Excluding children with Shiga toxin-producing Escherichia coli (STEC) from childcare until microbiologically clear of the pathogen, disrupts families, education, and earnings. Since PCR introduction, non-O157 STEC serotype detections in England have increased. We examined shedding duration by serotype and transmission risk, to guide exclusion advice. We investigated STEC cases aged <6 years, residing in England and attending childcare, with diarrhoea onset or sample date from 31 March 2018 to 30 March 2022. Duration of shedding was the interval between date of onset or date first positive specimen and earliest available negative specimen date. Transmission risk was estimated from proportions with secondary cases in settings attended by infectious cases. There were 367 cases (STEC O157 n = 243, 66.2%; STEC non-O157 n = 124, 33.8%). Median shedding duration was 32 days (IQR 20–44) with no significant difference between O157 and non-O157; 2% (n = 6) of cases shed for ≥100 days. Duration of shedding was reduced by 17% (95% CI 4–29) among cases reporting bloody diarrhoea. Sixteen settings underwent screening; four had secondary cases (close contacts’ secondary transmission rate = 13%). Shedding duration estimates were consistent with previous studies (median 31 days, IQR 17–41). Findings do not warrant guidance changes regarding exclusion and supervised return of prolonged shedders, despite serotype changes.