Background: Limited ankle dorsiflexion has been implicated as a contributing factor to plantar ulceration of the forefoot in diabetes mellitus. The purpose of this study was to compare outcomes for patients with diabetes mellitus and a neuropathic plantar ulcer treated with a total-contact cast with and without an Achilles tendon lengthening. Our primary hypothesis was that the Achilles tendon lengthening would lead to a lower rate of ulcer recurrence. Methods: Sixty-four subjects were randomized into two treatment groups, immobilization in a total-contact cast alone or combined with percutaneous Achilles tendon lengthening, with measurements made before and after treatment, at the seven-month follow-up examination, and at the final follow-up evaluation (a mean [and standard deviation] of 2.1 ± 0.7 years after initial healing). There were thirty-three subjects in the total-contact cast group and thirty-one subjects in the Achilles tendon lengthening group. There were no significant differences in age, body-mass index, or duration of diabetes between the groups. Outcome measures were time to healing of the ulcer, ulcer recurrence rate, range of dorsiflexion of the ankle, peak torque (strength) of the plantar flexor muscles, and peak plantar pressures on the forefoot. Results: Twenty-nine (88%) of thirty-three ulcers in the total-contact cast group and all thirty ulcers (100%) in the Achilles tendon lengthening group healed after a mean duration (and standard deviation) of 41 ± 28 days and 58 ± 47 days, respectively (p > 0.05). (One patient in the Achilles tendon lengthening group died before treatment was completed.) In the first seven months of follow-up, sixteen (59%) of the twenty-seven patients in the total-contact cast group who were available for follow-up and four (15%) of the twenty-seven patients in the Achilles tendon lengthening group who were available for follow-up had an ulcer recurrence (p = 0.001). At the time of the two-year follow-up, twenty-one (81%) of the twenty-six patients in the total-contact cast group and ten (38%) of the twenty-six patients in the Achilles tendon lengthening group had ulcer recurrence (p = 0.002). Compared with the group treated with the total-contact cast, the group treated with Achilles tendon lengthening had increased dorsiflexion and it remained increased at seven months (p < 0.001). Plantar flexor peak torque also decreased after Achilles tendon lengthening (p < 0.004), but it returned to baseline after seven months. Peak plantar pressures on the forefoot during barefoot walking were reduced (p < 0.0002) following Achilles tendon lengthening yet returned to baseline values within seven months after treatment. Conclusions: All ulcers healed in the Achilles tendon lengthening group, and the risk for ulcer recurrence was 75% less at seven months and 52% less at two years than that in the total-contact cast group. Achilles tendon lengthening should be considered an effective strategy to reduce recurrence of neuropathic ulceration of the plantar aspect of the forefoot in patients with diabetes mellitus and limited ankle dorsiflexion (≤5°). Level of Evidence: Therapeutic study, Level I-1a (randomized controlled trial [significant difference]). See Instructions to Authors for a complete description of levels of evidence.