This study tests whether the effectiveness of implementation intention-based interventions to increase fruit and vegetable intake in a young adult population can be enhanced using additional pre-intervention instructions and alternative formats; namely, an 'if-then' implementation intention versus a more general, 'global' plan that does not explicitly link a situational cue with a goal-directed response. Participants (N = 557) completed pre-test measures of planning, motivation and behaviour with respect to increasing their fruit and vegetable intake before being randomised to a 3 (intervention format: control vs. 'if-then' vs. 'global' implementation intentions) x 2 (pre-intervention instruction vs. no pre-intervention instruction) between-participants design. Results revealed a significant intervention format by time interaction, such that intake significantly increased by 0.50 portions in the if-then format condition compared with 0.31 in the global format and 0.01 in the control condition. These results suggest that 'if-then' manipulations are superior in promoting behaviour change in an applied setting. The use of pre-intervention instructions had no additional effect on behaviour, providing evidence for the efficacy of implementation intentions even when experimenter demand is reduced. Evidence is also presented to suggest that reported increases in intake are not related to demand characteristics.