We investigated the feasibility of rapid imaging of wrist trauma following casualty presentation and any subsequent effect on image quality and interpretability. All patients referred for wrist imaging were injected with 370 MBq 99Tcm-hydroxymethyl diphosphonate (HDP) and imaged 1, 2 and 3 h later. Palmar images were acquired on a 256 x 256 x 16 matrix using a high-resolution collimator, 140 keV photopeak and a 20% window. The images were scored qualitatively by four qualified observers in three categories: image quality, lesion detection and lesion localization. Statistical analysis indicated a significant improvement in scan quality with time, the mean difference (+/- standard error of the mean) between the 1 and 3 h scans being 0.81 +/- 0.07 (P = 0.001). No significant differences were seen in lesion detection (0.05 +/- 0.08; P = 0.51) or localization (0.14 +/- 0.08; P = 0.10). We conclude that imaging of wrist trauma is possible as early as 1 h post-injection of 99Tcm-HDP, although negative studies at 1 h require a 3 h image to maintain diagnostic accuracy.