Introduction Gas-based cryotherapy is the conventional ablative treatment for cervical pre-cancer in low-income settings, but the use of gas poses significant challenges. We compared the depth of necrosis induced by gas-based cryotherapy with two gas-free alternatives: cryotherapy using CryoPen,and thermoablation. Methods We conducted a five-arm randomized non-inferiority trial: double-freeze carbon dioxide (CO 2) cryotherapy (referent), single-freeze CO 2 cryotherapy, double-freeze CryoPen, single-freeze CryoPen, and thermoablation. Subjects were 130 women scheduled for hysterectomy for indications other than cervical pathology, and thus with healthy cervical tissue available for histological evaluation of depth of necrosis post-surgery. The null hypothesis was rejected (ie, conclude non-inferiority) if the upper bound of the 90% confidence interval (90% CI) for the difference in mean depth of necrosis (referent minus each experimental method) was <1.14 mm. Patient pain during treatment was reported on a scale of 0 (no pain) to 10 (worst pain). Results A total of 133 patients were enrolled in the study. The slides from three women were deemed unreadable. One patient was excluded because her hysterectomy was postponed for reasons unrelated to the study, and two patients were excluded because treatment application did not follow the established protocol. For the remaining 127 women, mean depth of necrosis for double-freeze CO 2 (referent) was 6.0±1.6 mm. Differences between this and other methods were: single-freeze CO 2 = 0.4 mm (90% CI-0.4 to 1.2 mm), double-freeze CryoPen= 0.7 mm (90% CI 0.04 to 1.4 mm), single-freeze CryoPen= 0.5 mm (90% CI-0.2 to 1.2 mm), and thermoablation = 2.6 mm (90% CI 2.0 to 3.1 mm). Mean pain levels were 2.2±1.0 (double-freeze CO 2 cryotherapy), 1.8±0.8 (single-freeze CO 2 cryotherapy), 2.5±1.4 (double-freeze CryoPen), 2.6±1.4 (single-freeze CryoPen), and 4.1±2.3 (thermoablation). Discussion Compared with the referent, non-inferiority could not be concluded for other methods. Mean pain scores were low for all treatments. Depth of necrosis is a surrogate for treatment efficacy, but a randomized clinical trial is necessary to establish true cure rates.