Streamflow data measured at hydrometric stations are affected by uncertainty due to water level (or stage) measurement errors. This uncertainty increases as the sensitivity of the stage-discharge controls decreases. A recently proposed method is used to demonstrate the role of control sensitivity in propagating stage uncertainty to streamflow uncertainty among other error sources. The method is first applied to a fictitious hydrometric station with five alternative scenarios of controls with variable sensitivity. The fictitious controls are designed on a flat weir without or with triangular or rectangular notches. The method is also evaluated using a real hydrometric station where the initially flat weir was changed to a weir with a rectangular notch. Both applications confirm the importance of the sensitivity of the controls, showing that a narrow notch in a flat control substantially reduces the uncertainty of daily, monthly and yearly discharges as well as the uncertainty of a low flow index. Quantifying such uncertainty component is essential to optimize the design of hydrometric stations and to provide streamflow data with acceptable uncertainty, especially for low flows.