Damming rivers changes sediment and nutrient cycles downstream of a dam in many direct and indirect ways. The Iron Gates I reservoir on the Yugoslavian-Romanian border is the largest impoundment by volume on the Danube River holding 3.2 billion m3 of water. Silica retention within the reservoir in the form of diatom frustules was postulated to be as high as 600 kt year−1 in previous studies using indirect methods. This amount of dissolved silicate was not delivered to the coastal Black Sea, and presumably caused a shift in the phytoplankton community there, and subsequent drastic decline in fishery. We directly quantified the amount of dissolved silicate (DSi) entering and leaving the reservoir for 11 continuous months. The budget based on these data reveals two important facts: (1) only about 4% of incoming DSi was retained in the reservoir; (2) the DSi concentrations were relatively low in the rivers upstream of the reservoir compared to regional and global averages. Thus damming the Danube at the Iron Gates could not have caused the decline in DSi concentrations documented downstream of the impoundment. Rather, this change in DSi must have occurred in the headwaters of the Danube River. Potential reasons include the construction of many dams upstream of the Iron Gates, hydrologic changes resulting in lower groundwater levels, and clogging of the riverbed limiting groundwater–river exchange.