Background and purpose: Brachytherapy is sometimes suggested as an adjuvant treatment after surgery of some tumours. When introducing this, it would be useful to have an estimate of the dose distribution to different body sites, both near and distant to target, comparing conventional external irradiation to brachytherapy. The aim of the present study was to determine radiation doses with both methods at different body sites, near and distant to target, in an experimental situation on an operated left sided breast cancer on a female Alderson phantom.Methods: Five external beam treatments with isocentric tangential fields were given by a linear accelerator. A specified dose of 1.0 Gy was given to the whole left sided breast volume. Five interstitial brachytherapy treatments were given to the upper, lateral quadrant of the left breast by a two plane, 10 needles implant. A dose of 1.0 Gy specified according to the Paris system was administered by a pulsed dose rate afterloading machine. Absorbed dose in different fixed dose points were measured by thermoluminescence dosimeters.Results: Both methods yielded an absorbed dose of the same size to the bone marrow and internal organs distant to target, 1.0–1.4% of the prescribed dose. There was a trend of lower doses to the lower half of the trunk and higher doses to the upper half of the trunk, respectively, by brachytherapy. A 90% reduction of absorbed dose with brachytherapy compared to external irradiation was found in the near-target region within 5 cm from target boundary where parts of the left lung and the heart are situated. If an adjuvant dose of 50 Gy is given with the external radiotherapy and brachytherapy, the absorbed dose in a part of the myocardium could be reduced from 31.8 to 2.1 Gy.Conclusions: Near target, brachytherapy yielded a considerably lower absorbed dose which is of special importance when considering radiation effects on the myocard and lungs. We could not demonstrate any difference of importance, in absorbed dose to dose points distant to target.