Brachytherapy dose distributions can be optimised by modulation of source dwell times. In this study dose optimisation in single planar interstitial implants was evaluated in order to quantify the potential benefit in patients.In 14 patients, treated for recurrent rectal and cervical cancer, flexible catheters were sutured intra-operatively to the tumour bed in areas with compromised surgical margin. Both non-optimised, geometrically and graphically optimised CT -based dose plans were made. The overdose index (OI), homogeneity index (HI), conformal index (COIN), minimum target dose, and high dose volumes were evaluated. The dependence of OI, HI, and COIN on target volume and implant regularity was evaluated. In addition, 12 theoretical implant configurations were analyzed.Geometrical and graphical optimisation improved the dose plans significantly with graphical optimisation being superior. Graphically optimised dose plans showed a significant decrease of 18%+/-9% in high dose volume (p<0.001). HI, COIN, and OI were significantly improved from 0.50+/-0.05 to 0.60+/-0.05, from 0.65+/-0.04 to 0.71+/-0.04, and from 0.19+/-0.03 to 0.15+/-0.03, respectively (p<0.001 for all). Moreover, minimum target dose increased significantly from 71%+/-5% to 80%+/-5% (p<0.001). The improvement in OI and HI obtained by optimisation depended on the regularity of the implant, such that the benefit of optimisation was larger for irregular implants. OI and HI correlated strongly with target volume limiting the usability of these parameters for comparison of dose plans between patients.Dwell time optimisation significantly improved the dose distribution regarding homogeneity, conformity, minimum target dose, and size of high dose volumes. Graphical optimisation is fast, reproducible and superior to geometric optimisation.