The data from Meisberger et al. [Radiology 90, 953-957 (1968)] are often used as a basis for dose calculations in brachytherapy. In order to describe the absorbed dose in water around a brachytherapy point source, Meisberger provided a polynomial fit for different isotopes taking into account the effect of attenuation and scattering. The validity of the Meisberger coefficients is restricted to distances up to 10 cm from the source, which is regarded to be satisfactory for most brachytherapy applications. However, for more distant organs it may lead to errors in calculated absorbed dose. For this reason dose measurements have been performed in air and in water around a high activity 60Co source used in high dose rate brachytherapy. Measurements were carried out to distances of 20 cm, using ionization chambers. These data show that at a distance of about 15 cm the amount of scattered radiation virtually equals the amount of primary radiation. This emphasizes the contribution of scattered radiation to the dose in healthy tissue far from the target volume, even with relatively high energy photon radiation of 60Co. It is also shown that the Meisberger data as well as the approach of Van Kleffens and Star [Int. J. Radiat. Oncol. Phys. 5, 557-563 (1979)] lead to significant errors in absorbed dose between distances of 10 and 20 cm from the source. In addition to these measurements, the Monte Carlo code has been used to calculate separately primary dose and scattered dose from a cobalt point source. The calculated results agree with the experimental data within 1% for a most distant dose scoring region.