The primary objective of this work was to estimate the absorbed dose distribution to the bone marrow of six multiple myeloma patients who received holmium-166 (166Ho) DOTMP (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetramethylene-phosphonic acid) for the purpose of bone marrow ablation. A methodology based on gamma camera images was developed to estimate the regional absorbed dose distributions delivered to the bone marrow, and this was compared with values calculated from the MIRD technique and bone marrow biopsies. The activity concentration in various skeletal regions was calculated from the activity in the region of interest (ROI) drawn on whole body gamma camera images, and the mass of bone in each ROI was derived from a dual x-ray absorptiometry image. The radiation absorbed dose to the bone marrow was calculated from this activity concentration using an adaptation of Bragg-Gray cavity theory. The radiation absorbed dose delivered to the bone marrow in the six patients calculated from the MIRD "S" factors ranged from 15.0 to 46.3 Gy. The gamma camera measured activity concentration in skeletal regions predominantly composed of trabecular bone was approximately five to six times higher than that in cortical regions. The skeletal activity concentration in each patient ranged from highs in such regions as the ribs to lows in such regions as the shin and foot by a factor of nearly 20, producing a heterogeneous distribution of absorbed dose within the marrow. Dose volume histograms calculated for these patients indicated that 15%-20% of the marrow received an absorbed dose significantly larger than the average value, while 5%-10% of the marrow received a substantially lower dose. Weighted mean dose estimates from the regional technique were typically 30% greater than the average dose estimates calculated with the MIRD "S" factors. Finally, absorbed dose estimates for the marrow calculated from the regional technique correlated more closely with the clinical response of blood cells and abnormal proteins measured in bone marrow aspirates and peripheral blood samples than estimates from the MIRD "S" factors.