This case of a middle aged woman with IL-12p40 deficiency and recurrent Mycobacterium kansasii infection demonstrates that in vitro sensitivities do not reliably predict the clinical response of non-tuberculous mycobacteria (NTMB), and emphasises the need to exercise caution when reducing treatment intensity in patients with type 1 cytokine defects on the basis of in vitro minimum inhibitory concentration (MIC) data alone.