Purpose: To quantify the extent of additional source placement needed to perfect an implant after execution by standard techniques, assuming that uniform 5 mm treatment margins (TMs) is the criteria for perfection. Materials and Methods: Ten consecutive, unselected patients treated with I-125 brachytherapy were studied. Source placement is planned just inside or outside of the prostatic margin, to achieve a minimum 5 mm TM and a central dose of 150%–200% of the prescription dose. The preimplant prostate volumes ranged from 24 to 85 cc (median: 35 cc). The number of sources implanted ranged from 48 to 102 (median: 63). Axial CT images were acquired within 2 h postoperatively for postimplant dosimetry. After completion of standard dosimetric calculations, the TMs were measured and tabulated at 45° intervals around the prostate periphery at 0.0, 1.0, 2.0, and 3.0 cm planes. Sources were then added to the periphery to bring the TMs to a minimum of 5 mm at each measured TM, resulting in a modified implant. All margin modifications were done manually, without the aid of automated software. Results: Patients’ original (unmodified) D90s ranged from 111% to 154%, with a median of 116%. The original V100s ranged from 94% to 99%, with a median of 96%. No patient required placement of additional sources to meet a minimum D90 of 90% or a minimum V100 of 80%. In contrast, patients required from 7 to 17 additional sources (median: 11) to achieve minimum 5 mm TMs around the entire prostatic periphery. Additional sources equaled from 12% to 24% of the initial number of sources placed (median: 17%). By adding sufficient peripheral sources to bring the TMs to a minimum 5 mm, patients’ average V100 increased from 96% to 100%, and the average D90 increased from 124% to 160% of prescription dose. In the course of achieving a minimum 5 mm TM, the average treatment margin for all patients combined increased from 5.5 to 9.9 mm. The number of sources needed to bring the TMs to a minimum 5 mm was loosely correlated with the preimplant prostate volume and the change in prostate volume from implant-related swelling. Adding sufficient sources to achieve minimum 5 mm TMs increased the prostate volume receiving greater than 200% of the prescription dose (V200) from 39% to 58%, and increased the average urethral point dose (2.00 cm inferior to the bladder) from 154% to 171% of the 144 Gy prescription isodose. Conclusions: Minimum 5 mm TMs are not uniformly achieved with current implant techniques. It seems that doing so, even in experienced hands, will require a reappraisal of our implant techniques, or the addition of intraoperative dosimetric analysis with the capacity to substantially modify the implant with extra sources. © 2002 American Association of Physicists in Medicine.