Germane to the production and use of health education advertisements is their effectiveness. MacLachlan, Carr, Fardell, Maffesoni, and Cunningham (1997) proposed that an HIV/AIDS health education message is more effective when the ego state “hooked” by the advertisement is complementary (parallel) to the ego state experienced during sexual interaction. The aim of this study was to test this model further and, through minor modification of the original design, to gather additional salient subsidiary data. One hundred and forty-four sexually active Australian undergraduates viewed two government-sponsored television advertisements (“Reaper” and “Beds”) designed to prevent the spread of acquired immune deficiency syndrome (AIDS). After the viewing, the participants were asked to identify the ego states projected by the advertisements, the ego states they would have preferred to have projected, and their prediction of their own ego state during sexual interaction with both a steady and a casual partner. The participants rated the effectiveness of each of the two advertisements on a categorical scale (poor, satisfactory, good, and excellent) for each of two dimensions (conveying knowledge about HIV/ AIDS and conveying the need for safe sexual practices). Higher ratings for effectiveness on both dimensions were found for parallel transactions represented as matches between projected ego states and preferred ego states at the time of viewing (x 2 (N = 143) = 33.7 p < .001; x 2 (3, N = 143) = 43.1 p < .001; x 2 (3, N = 140) = 36.1 p < .001; and x 2 (N = 141) = 28.1 p < .001), but not for matches (parallel transactions) between advertisement projected ego states and predicted ego states for self during sexual interactions. Analysis of subsidiary data revealed that predicted ego states did not differ according to type of sexual partner (steady, casual); both advertisements were rated as more effective in conveying the need for safe sexual practices than knowledge about HIV/AIDS (z − 6.8, p < .001 & z − 8.9 p < .001); the “Beds” advertisement was rated more effective than the “Reaper” advertisement for both knowledge impact and need for safe practices (z − 2.67, p < .01 & z − 6.0 p < .001); and the majority (63% and 66%) of respondents selected Adult to Adult as their preferred ego states for the advertisements to project. The utility of transactional analysis in the social marketing and evaluation of health education is discussed with reference to overcoming its limitations through the development of a transactional positioning approach.