People respond differently when information is framed either positively or negatively (frame valence). Two prominent models propose that the effects of valence are moderated by (1) the method of framing (attributes vs. goals: Levin, Schneider, & Gaeth, 1998) and (2) perceived risk (Rothman & Salovey, 1997). This experiment (N=200) explores the joint influence of both of these moderators with respect to decisions about a flu vaccination. The study extends previous work by integrating these two models and exploring the moderating effects of two different aspects of perceived risk (personal outcome effectiveness and procedural risk). The results show that personal outcome effectiveness indirectly links frames to intentions. Procedural risk moderates the relationship between valence and method in a manner consistent with predictions from Levin et al.. Partial support for the model proposed by Rothman and Salovey are observed for goal frames only.