This exploratory randomised controlled trial examined the effectiveness of a novel short messaging service intervention underpinned by the theory of planned behaviour (TPB) in improving insulin administration in young adults with type 1 diabetes and the role of moderating variables. Those in the intervention condition (N = 8) received one daily text message underpinned by TPB constructs: Attitudes, subjective norms, perceived behavioural control and intention. Those in the control condition (N = 10) received weekly general health messages. Self-reported insulin administration was the main outcome measure; conscientiousness and consideration of future consequences (CFC) were measured as potential moderators. Analyses of covariance revealed no main effects of condition for morning and afternoon injections but a marginally significant effect for evening injections (p = .08). This main effect was qualified by significant interactions of condition with conscientiousness (p = .001), CFC (p = .007) and a three-way interaction among condition, conscientiousness and CFC (p = .009). Exploration of the interactions indicated the intervention significantly improved evening injection rates only in the low conscientiousness and low CFC groups. This effect was particularly strong among those low in both conscientiousness and CFC. Further investigation is warranted, using more objective measures of insulin adherence in a larger sample.