The present work has been carried out as a support for site selection for The European Solar Telescope (EST), a 4 m telescope project that will be located in the Canary Islands Observatories. For such purpose, we present a comparative analysis of a long time-series of 17 yr of climate parameters with significance in astronomical site testing for Teide (OT) and Roque de Los Muchachos (ORM) observatories. The climate parameters have been obtained from a climate reanalyses database with a spatial resolution of 11 km × 11 km, which has been compared with local averaged data using Automatic Weather Stations (AWS) and radiosounding. The results show the reliability of using climate reanalysis databases for site-testing both at the heights of the OT and ORM observatories (boundary layer) and in the free atmosphere. Climate parameters analyzed are relative humidity, temperature, wind velocity, and wind direction. For the boundary layer, we have obtained full time-series, annual behavior, and statistical distributions for each parameter. Monthly averaged values are given by using a 17 yr time period. Statistics are given for two time periods: Summer (June to August) and non-summer (the rest). The summer period has been selected according to the most intense observational period along the year for the EST. Day/night variations of the climate parameters have been also analyzed. Both observatories show values in the range 6%-70% for relative humidity, 2.6°C-19°C for temperature, and 5-12 m s-1 for wind velocity, 95% of the time; median values are around 25%, 10°C, and 5 m s-1, respectively. The most frequent wind direction (wind blowing from) is West for both observatories (~20%), with a second peak at Northeast (~19%) for ORM and Northwest for OT (~17%). Less frequent wind directions are South and Southeast (≲5% in both cases). No significant difference between OT and ORM in these climate parameters have been found. Using the relative humidity, temperature, wind, and velocity limits for EST we have estimated an operational time ≳98%. Our results confirm the quality of ORM and OT observatories with stable and predictable weather conditions.