Atmospheric water vapor pressure is an essential meteorological control on land surface and hydrologic processes. As it is not as frequently observed as other meteorologic conditions, it is often inferred through the August–Roche–Magnus formula by simply assuming dew point and daily minimum temperatures are equivalent or by empirically correlating the two temperatures using an aridity correction. The performance of both methods varies considerably across different regions and during different time periods; obtaining consistently accurate estimates across space and time remains a great challenge. Here, an interpretable Long Short-Term Memory (iLSTM) network conditioned on static, location specific attributes is proposed to estimate the daily vapor pressure. This approach allows for training a single transferable model using ensemble data from multiple sites and exploring the quantitative dependency of vapor pressure prediction on multiple environmental variables and their histories. To evaluate this approach, three iLSTM model configurations were developed, each considering different site attributes as static variables. For each configuration, multiple model realizations were trained using 83 FLUXNET sites in the United States and Canada, where each realization corresponds to different withheld groups of sites used for model evaluation. Results show that the iLSTM networks noticeably improve the estimation accuracy in comparison with the two assumption-based methods for most sites, reducing the failure rate from 32 % to 10.9 % for the best iLSTM model configuration. Additionally, this network provides reasonable insights into both the relative importance of the time-series input variables and their temporal importance. This method is found to be effective for imputing vapor pressure across space and time.