Context: The inconclusive evidence regarding long-term safety of recombinant human growth hormone (rhGH) therapy underlines the need for long-term large-scale cohorts. Objective: To assess long-term mortality and cancer incidence among patients treated with rhGH during childhood in Israel. Design: A population-based cohort study. Setting: Data were retrieved from a national register established in 1988. Mortality data from the national population register were available through 31 December 2014. Data on cancer incidence from the national cancer registry were available through 31 December 2012. Participants: All patients ≤19 years approved for rhGH treatment during 1988–2009 were included. Patients were assigned to three risk categories, according to the underlying condition leading to growth disorder. Main outcome measures: All-cause mortality and cancer incidence rates were calculated, based on person-years at risk. Standardized mortality ratios (SMRs) and standardized incidence ratios (SIRs) were calculated, using the Israeli general population as a reference. Results: Included were 1687 patients assigned to the low-risk category and 440 patients assigned to the intermediate-risk category. In the low-risk category, all-cause mortality and cancer incidence were not significantly different than expected (SMR 0·81, 95% CI 0·22–2·08 and SIR 0·76, 95% CI 0·09–2·73). In the intermediate-risk category, all-cause mortality and cancer incidence were significantly higher than expected (SMR 4·05, 95% CI 1·62–8·34 and SIR 4·52, 95% CI 1·22–11·57). Conclusions: No increased risk of mortality or cancer incidence was found in low-risk patients treated with rhGH during childhood. Patients with prior risk factors were at higher risk of both mortality and cancer.