AIMS: To analyse the cost effectiveness of using the moxibustion technique to correct non-vertex presentation and to reduce the number of caesarean sections performed at term. METHOD: A deterministic model of decision analysis has been developed to analyse the cost of treatment in which heat is applied by moxibustion (the combustion of Artemisia vulgaris) at acupuncture point BL67 for pregnant women with non-vertex fetal position at 33-35?weeks' gestation. This approach was compared with conventional treatment recommendations based on the knee-chest posture technique. The costs were obtained mainly from data provided by the Andalusian Public Health System. Effectiveness data for the baseline analysis were taken from a previous clinical study. A secondary analysis was performed based on a meta-analysis conducted using random effects analysis, by reference to studies published in recent systematic reviews of moxibustion versus conventional treatment, in order to make the results generalisable to other healthcare settings. Deterministic and probabilistic sensitivity analyses were performed under diverse assumptions to assess the uncertainty of the result. RESULTS: The baseline analysis shows that the application of moxibustion prevents 8.92% of deliveries with non-vertex presentation compared with conventional treatment, with an average cost saving of 107.11 per delivery, mainly due to the cost saving from avoiding the need for caesarean section. The meta-analysis revealed a relative risk of the version of non-vertex presentation at term of 0.34 (95% CI 0.16 to 0.76). The sensitivity analysis showed that moxibustion can avoid 0.34 caesarean sections, with an incremental cost per delivery ranging from 68 to -640 for moxibustion versus conventional treatment. CONCLUSIONS: Moxibustion treatment applied at acupuncture point BL67 can avoid the need for caesarean section and achieve cost savings for the healthcare system in comparison with conventional treatment.