Adaptive radiotherapy involves altering the treatment plan according to variations in patient anatomy and set-up. This relies upon an accurate representation of the changing dose distribution within the patient, requiring a full dose recalculation. This work proposes a novel workflow using the planned dose distribution to assess dose coverage in three-dimensional verification CT studies acquired at the time of treatment delivery, using an overlay technique, in lieu of a recalculated dose distribution. The concept has been validated in a pilot study of 10 patients, each with 7-10 on-treatment CT studies. Differences between the geometric shape of the treatment plans for the 95% isodose and the 95% isodose obtained when the planned geometry was recalculated from the verification CT dataset were quantified. Dosimetric coverage of the verification clinical target volume (vCTV) was assessed for both the proposed overlay technique and the recalculated "delivered" dose distribution, and the conclusions on adequacy were compared. Results were consistent with geometric uncertainties of the dose calculation matrix (5 x 5 x 5 mm), suggesting that differences in the geometric shape of the 95% isodose are not significant for normal variations in patients' anatomy. Decisions on adequacy of vCTV coverage were consistent in 80 out of 87 cases, with discrepancies limited to a maximum of three axial slices per study within the range 0.5-4.5 mm (mean, 1.6 mm). The proposed dosimetric overlay technique has been validated and found to be an acceptable method of image-guided radiotherapy of the prostate suitable for effective implementation in the treatment clinic.