OBJECTIVES:  To determine the most cost-effective sequence of interventions for the management of refractory primary hyperhidrosis in secondary care in the UK NHS, and to establish the expected value of undertaking additional research on clinical outcome and utility parameters. METHODS:  A de novo probabilistic model was developed, which modelled sixty-four different treatment sequences for axillary hyperhidrosis. Treatments included: aluminium chloride, medication(M), botulinum toxin(BTX), iontophoresis(I) sponge, curettage(C) and endoscopic thoracic sympathectomy(E). A 48-year time horizon was used, with health outcomes expressed as QALYs. The perspective was that of the UK NHS and Personal Social Services. Sensitivity analyses were conducted, alongside both the population expected value of perfect- and partial perfect-information (EVPPI) for model parameters. These were calculated for different willingness to pay thresholds (WTP), incidence rates and effectiveness of treatments. Costs were calculated in GBP (£) for the 2015 price year. RESULTS:  Base-case results indicated that IBMCE was the most cost-effective sequence; ICER of £9,304/QALY; probability cost-effective 0.8. The next most cost-effective sequences involved medication moving before BTX or disappearing from the sequence. This reflects the uncertainty in the effectiveness estimate of medication vs placebo. Most sensitivity and scenario analyses had little effect on the results. The population EVPPI for medication vs placebo and for curettage vs BTX effectiveness was > £3 million for all scenarios; unless the incidence of axillary hyperhidrosis was 0.5% and WTP was £20,000/QALY. CONCLUSIONS:  The analysis found that the treatment sequence IBMCE was most cost-effective. Despite the limited evidence, iontophoresis and BTX maintained their position as first- and second-line treatment options in the majority of cost-effective sequences. The value of information analysis suggested that whilst further research of medication vs placebo may be of value, with medication either coming before BTX in the sequence or dropping out altogether.