The article introduces the GMTKN55 benchmark database, an advanced and improved version of the GMTKN30 database, designed for assessing general main group thermochemistry, kinetics, and noncovalent interactions. The new database includes 55 benchmark sets, a significant increase from the 30 sets in GMTKN30, and provides higher-quality reference values for most sets. The authors emphasize the importance of better reference values and the need for London-dispersion corrections in density functional theory (DFT) treatments of thermochemical problems. They evaluated 217 variations of dispersion-corrected and -uncorrected DFT approximations, focusing on 83 of them to identify robust and reliable approaches. Double-hybrid functionals are highlighted as the most reliable for thermochemistry and noncovalent interactions, with specific recommendations for DFT methods. The article also discusses the limitations of popular methods like B3LYP and encourages method developers to use GMTKN55 for cross-validation studies. The detailed description of each benchmark set and the evaluation of DFT methods are provided, along with recommendations for future applications.The article introduces the GMTKN55 benchmark database, an advanced and improved version of the GMTKN30 database, designed for assessing general main group thermochemistry, kinetics, and noncovalent interactions. The new database includes 55 benchmark sets, a significant increase from the 30 sets in GMTKN30, and provides higher-quality reference values for most sets. The authors emphasize the importance of better reference values and the need for London-dispersion corrections in density functional theory (DFT) treatments of thermochemical problems. They evaluated 217 variations of dispersion-corrected and -uncorrected DFT approximations, focusing on 83 of them to identify robust and reliable approaches. Double-hybrid functionals are highlighted as the most reliable for thermochemistry and noncovalent interactions, with specific recommendations for DFT methods. The article also discusses the limitations of popular methods like B3LYP and encourages method developers to use GMTKN55 for cross-validation studies. The detailed description of each benchmark set and the evaluation of DFT methods are provided, along with recommendations for future applications.