Perceptions of risk for health outcomes are integral to many theories of health behaviour, and are often targeted in interventions. Evidence suggests that affective responses to risk, including worry, are empirically distinguishable from commonly used perceived risk measures such as perceived susceptibility. The aims of this meta-analysis were to (1) examine if perceived susceptibility and worry can be independently influenced, and what manipulation types are most effective at changing each construct and (2) examine the efficacy of interventions to change worry and perceived susceptibility. Thirty-eight studies using 43 separate samples provided 78 independent comparisons that were meta-analysed using the inverse variance method with random-effects modelling. The overall effect size (d) was 0.50, 95% CI [0.362, 0.632] for perceived susceptibility; and 0.25, 95% CI [0.148, 0.349] for worry. Effect sizes for perceived susceptibility were significantly related to those for worry, B=0.495, p < 0.001. Moderators of these effects are discussed. The present meta-analysis provides further evidence that perceived susceptibility and worry are distinguishable but related constructs, and that it is possible to perturb one and not the other.