This paper presents a critical compilation of accurate stellar masses and radii from 95 detached binary systems containing 190 stars, with both masses and radii known to within ±3%. These systems are non-interacting, allowing stars to evolve as single stars. The sample extends the mass range and includes an extragalactic binary, and all data are reanalyzed with consistent assumptions and physical constants. Additional parameters like effective temperature, metal abundance, and rotational velocity are included. The data reveal stellar evolution effects clearly and are used to test stellar evolution models. Significant structural differences between fast-rotating, active stars and single stars are confirmed. The data also allow testing of convection, diffusion, and other non-classical effects in stellar models. The data enable empirical calibrations for single stars above 0.6 M☉, with errors of 6% and 3% for mass and radius, respectively. These calibrations agree with independent determinations for host stars of transiting extrasolar planets and with stellar models constrained by trigonometric parallaxes and spectroscopic values. The data also allow analysis of tidal evolution, testing rotational synchronisation and orbital circularisation. Pseudo-synchronisation formulas predict observed rotations well, except for young or widely-separated stars. General Relativity predictions for apsidal motion agree with observations. The data also allow calibrations for single stars with reliable T_eff, log g, and [Fe/H]. Long-baseline interferometry offers promise for future mass and radius determinations. A list of 23 interferometric binaries with masses known to better than 3% is provided. The paper emphasizes that the data are of high quality but not complete or unbiased, making them unsuitable for statistical analysis. The selection criteria require that stars have evolved as single stars and that masses and radii are accurate to better than 3%. The data include systems with detectable companions and eclipsing binaries. The analysis techniques used include high-quality data, appropriate methods, and critical error assessment. The physical constants used in mass and radius calculations are discussed, with recommendations for accurate values. The data allow analysis of light curves, effective temperatures, and other parameters. The data also include effective temperatures, metal abundances, rotational velocities, and interstellar reddening. The data are used to compute luminosities, distances, and ages. The data show significant effects of stellar evolution, with stars moving horizontally to cooler temperatures and diagonally towards the upper right in temperature-radius diagrams. The data are used to test stellar evolution models, with AI Phe as a key example. The data show that models with accurate parameters fit observations well, but uncertainties in metallicity and helium abundance affect results. The data also show that stars with convective envelopes evolve differently from those with radiative envelopes. The data allow analysis of tidal evolution, with circularisation and synchronisation effects observed in different systems. The data are used to study tidal circularisation andThis paper presents a critical compilation of accurate stellar masses and radii from 95 detached binary systems containing 190 stars, with both masses and radii known to within ±3%. These systems are non-interacting, allowing stars to evolve as single stars. The sample extends the mass range and includes an extragalactic binary, and all data are reanalyzed with consistent assumptions and physical constants. Additional parameters like effective temperature, metal abundance, and rotational velocity are included. The data reveal stellar evolution effects clearly and are used to test stellar evolution models. Significant structural differences between fast-rotating, active stars and single stars are confirmed. The data also allow testing of convection, diffusion, and other non-classical effects in stellar models. The data enable empirical calibrations for single stars above 0.6 M☉, with errors of 6% and 3% for mass and radius, respectively. These calibrations agree with independent determinations for host stars of transiting extrasolar planets and with stellar models constrained by trigonometric parallaxes and spectroscopic values. The data also allow analysis of tidal evolution, testing rotational synchronisation and orbital circularisation. Pseudo-synchronisation formulas predict observed rotations well, except for young or widely-separated stars. General Relativity predictions for apsidal motion agree with observations. The data also allow calibrations for single stars with reliable T_eff, log g, and [Fe/H]. Long-baseline interferometry offers promise for future mass and radius determinations. A list of 23 interferometric binaries with masses known to better than 3% is provided. The paper emphasizes that the data are of high quality but not complete or unbiased, making them unsuitable for statistical analysis. The selection criteria require that stars have evolved as single stars and that masses and radii are accurate to better than 3%. The data include systems with detectable companions and eclipsing binaries. The analysis techniques used include high-quality data, appropriate methods, and critical error assessment. The physical constants used in mass and radius calculations are discussed, with recommendations for accurate values. The data allow analysis of light curves, effective temperatures, and other parameters. The data also include effective temperatures, metal abundances, rotational velocities, and interstellar reddening. The data are used to compute luminosities, distances, and ages. The data show significant effects of stellar evolution, with stars moving horizontally to cooler temperatures and diagonally towards the upper right in temperature-radius diagrams. The data are used to test stellar evolution models, with AI Phe as a key example. The data show that models with accurate parameters fit observations well, but uncertainties in metallicity and helium abundance affect results. The data also show that stars with convective envelopes evolve differently from those with radiative envelopes. The data allow analysis of tidal evolution, with circularisation and synchronisation effects observed in different systems. The data are used to study tidal circularisation and