Gridded information on past and present snow cover is essential for climate services in snow-dominated regions like the Alps. As part of the SPASS project (SPAtial Snow climatology for Switzerland), we developed the first long-term gridded datasets of daily snow water equivalent (SWE) and snow depth, covering Switzerland at 1 km resolution since 1962. We describe our method for generating these datasets and their evaluation for climatological analyses. Two dataset families were produced: 1) Climatological dataset (since 1962): Bias-adjusted snow model results using a quantile-mapping method. 2) Assimilation-based dataset (since 1999): A higher-quality snow model that incorporates snow depth observations. Comparing both datasets shows good overall performance, particularly in bias and correlation. Errors remain acceptable, except for ephemeral snow and short time aggregations like weeks. Validation against in-situ station data for yearly, monthly, and weekly values at different elevations indicates only slightly better performance for the higher-quality dataset, confirming the robustness of the quantile-mapping method. A trend analysis of yearly mean snow depth from both station-based and gridded data revealed strong agreement in trend direction and significance across elevations. However, at low elevations, gridded datasets tend to overestimate the strength of decreasing trends. Overall, our results confirm that the new snow datasets perform well but may show the largest uncertainties at low elevations, single grid points, or short time periods. Despite some limitations, these high-resolution snow datasets offer valuable insights into long-term trends and climate variability, supporting applications such as anomaly maps and elevation-based trend analysis.