The transcription factor Nrf2 protects against a number of experimental pathologies, and is a promising therapeutic target. The clinical investigation of a potent Nrf2-inducing agent, the triterpenoid (TP) bardoxolone methyl (BARD), was recently halted due to adverse cardiovascular events in chronic kidney disease patients, although the underlying mechanisms are yet to be resolved. The majority of small molecule Nrf2 inducers are electrophilic and trigger Nrf2 accumulation via the chemical modification of its redox-sensitive repressor Keap1. Therefore, it is pertinent to question whether the therapeutic targeting of Nrf2 could be hindered in many cases by the inherent reactivity of a small molecule inducer towards unintended cellular targets, a key mechanism of drug toxicity. Using H4IIE-ARE8L hepatoma cells, we have examined the relationship between (a) Nrf2 induction potency, (b) toxicity and (c) in vitro therapeutic index (ratio of b:a) for BARD and a number of other small molecule activators of Nrf2. We show that BARD exhibits the highest potency towards Nrf2 and the largest in vitro therapeutic index amongst compounds that have been investigated clinically (namely BARD, sulforaphane and dimethylfumarate). Through further examination of structurally-related TPs, we demonstrate that an increase in potency towards Nrf2 is associated with a relatively smaller increase in toxicity, indicating that medicinal chemistry can be used to enhance the specificity of a compound as an inducer of Nrf2 signalling whilst simultaneously increasing its therapeutic index. These findings will inform the continuing design and development of drugs targeting Nrf2.