The current thesis delineates a programme of research that sought to design, develop, and refine an IRAP that could be used as a reasonably reliable and valid measure of implicit homonegativity. A novel feature of the programme was the introduction of a multi-dimensional approach for screening participant sexual orientation. An additional purpose of the current research was to provide the first systematic analysis of implicit homonegativity in Ireland. Over the course of a series of experiments, the IRAP was used to explore implicit homonegativity with a particular focus on: (a) its malleability as a result of situational/context manipulation effects (Experiment 1; Chapter 3) and prior exposure to gay-related exemplars (Experiment 2; Chapter 4); (b) known-group differences and the implications of sensitive multi-dimensional sexual orientation screening (Experiment 3 & 4; Chapters 5 & 6); (c) the impact of response latency restrictions (Experiment 4 & 6; Chapters 6 & 8); (d) the moderating impact of selfreported motivation to control homonegativity on IRAP responses (Experiment 1; Chapter 3); and (e) the impact of single versus multiple labels on the IRAP (Experiment 6; Chapter 8). In addition, Chapter 7 (Experiment 5) again using a known-groups approach, presents the first IAT study to investigate implicit homonegativity in Ireland. Support for the reliability of the IRAP was provided when the same general pattern of pro-straight and anti-gay biases were observed across most of the IRAPs (although moderated by a number of variables). The known-groups studies (Experiments 3 and 4) provided strong support for the validity of the Homonegativity-IRAP (and demonstrated the utility of the multi-dimensional approach to sexual orientation screening that was employed) because it clearly discriminated between Exclusive-Heterosexuals (EH), Non- Exclusive-Heterosexuals (NEH) and Gay, Lesbian, Bisexual (GLB) groups. In addition, reducing the response latency criterion from 3000ms to 2000ms served to increase the size of the IRAP effects and produce a pattern of responding that was more consistent with sexual orientation group status. Unlike the 3000ms IRAP and the IAT, the 2000ms IRAP did not correlate with a measure of social desirability. Arguably the most important finding was that subtle changes in the number and type of stimuli that were employed in the IRAP had a dramatic impact on the size of the D-IRAP effects. Overall, the research reported in the current thesis provides support for the reliability and validity of the Homonegativity-IRAP and suggests that it is a relatively robust measure that could usefully be employed in future investigations of implicit homonegativity.