ABSTRACT: The application of otolith tools to infer the movement history of fishes requires controlled experiments to validate methods and assess confidence in inferences gained for wild specimens. The Delta Smelt (Hypomesus transpacificus) is a critically endangered estuarine fish endemic to the San Francisco Estuary (SFE), California, United States, and serves as a key indicator species. Understanding variation in habitat use and migratory behaviors of this species is critical for developing effective conservation and management actions. Here, we conducted an experiment using known-age cultured Delta Smelt and mixtures of coastal seawater and freshwaters of the upper SFE to manipulate their salinity history and examine the temporal resolution and accuracy of salinity reconstructions using otolith strontium isotope ratios (⁸⁷Sr/⁸⁶Sr). Results indicated that salinity reconstructions from Delta Smelt otolith ⁸⁷Sr/⁸⁶Sr provide accurate salinity estimates (bias <0.5 ppt) in waters below 6 ppt when using an SFE-wide strontium isotope to salinity mixing model. Instantaneous transitions from fresh to brackish waters at 60 days-post-hatch (dph) with a salinity of 3 ppt could be detected within ~1 week (59 dph ±6, SD), while transitions to higher salinities of 6 ppt at 120 dph were more difficult to detect and uncertainty increased to about 3 weeks (132 dph ±24, SD). These results confirm the utility of otolith geochemistry for examining the movement behaviors of fishes in complex and dynamic estuarine environments where salinity gradients can vary rapidly over short time frames.