Abstract In an effort to determine accurate orbital and physical properties of a large number of bright stars, a method was developed to fit simultaneously stellar parameters (masses, luminosities, effective temperatures), distance, and orbits to the available data on multiple systems, namely the combined and differential photometry, positional measurements, radial velocities (RVs), accelerations, etc. The method is applied to a peculiar resolved triple system HIP 86286. The masses of its components estimated using observations and standard relations are 1.3, 0.9, and 0.9 M⊙; the main star is a G8IV subgiant, while its two companions are main-sequence dwarfs. The inner and outer orbital periods are 35 and 287 years, respectively, and the orbits are nearly coplanar. The second system, HIP 117258, is an accelerating star with a resolved companion; its 35.7-yr orbit based on relative astrometry and precise RVs yields the secondary mass of 0.95 M⊙, much larger than inferred from the photometry. The apparent paradox is explained by assuming that the secondary is a close pair of M-type dwarfs with yet unknown period.