The primary aim of this thesis was to investigate whether commonly used personality, anxiety and experiential avoidance trait related measures provide any predictive utility in identifying observed levels of Pavlovian conditioning and the symbolic or semantic generalisation of fear and avoidance. A small number of previous studies had already attempted to correlate empirically observed levels of generalised threat and avoidance responding with scores on a number of trait and experiential avoidance questionnaires but had limited success. However, these studies focused on generalisation along perceptual gradients, while this thesis focused more on ecologically valid symbolic and semantic generalisation. Seven exploratory computer-based experiments are outlined, six of which provided participants with the opportunity to successfully avoid the US and then subsequently generalise either SCRs, US expectancy ratings or instrumental avoidance responses across symbolically or semantically related nonsense or English words. Experiment 1 sought to address the previous omission of trait anxiety and experiential avoidance measures from the symbolic generalisation literature. The paradigm consisted of three phases; equivalence learning, fear and avoidance learning and finally, probes for generalisation. Results indicated that avoidance behaviour and threat-expectancy readily conditioned and then generalised to symbolically related stimuli. However, trait anxiety and experiential avoidance do not predict symbolic generalization of avoidance. Experiments 2a and 2b returned to the examination of less complex forms of fear and avoidance by comparing the relationship between trait scores and Pavlovian conditioning rates to that between trait measures and semantic generalisation rates. Specifically, Experiment 2a employed a Pavlovian conditioning method, with only a single phase of avoidance learning, while Experiment 2b included a generalisation probe phase, using real words and their synonyms as cues. Both experiments successfully demonstrated the ease with which avoidance learning and generalisation occurs, as well as identifying a number of tantalising co-relations between the trait questionnaires and the dependent measures. Experiment 3, 4, 5 and 6, all used the Boyle et al. (2016) paradigm, comprising of 3 phases; fear conditioning, avoidance conditioning and final probes, with a range of procedural modifications to attempt to identify specific effects. Experiment 3 produced successful conditioning of two cues across all phases. Generalisation between the cues was supported by discriminated differences in avoidance responding and US expectancy, but not for arousal response magnitudes. Similar to the previous experiment, the predictive utility of the questionnaires was more pronounced for the conditioned responses than for generalised ones. In an attempt to address a number of possible confounds, Experiment 4 replaced the single press low-cost avoidance response from Experiment 3, with a higher physical (20x press) cost response. Overall, regardless of participant’s US avoidance success, rates of attempted avoidance (i.e., ≥ 1 key-presses) to the CS+ and CS- during all phases supported the successful conditioning of safety and threat to the cues, which then was shown to semantically generalise. A participant’s success in regularly cancelling the delivery of the US, was also related to their likelihood of attempting avoidance during probe trials. Questionnaire scores were not significantly correlated with either the observed rates of generalisation or individual success in making an avoidance response. Experiment 5 sought to examine whether the introduction of a novel unrelated probe stimulus, during the final phase, would result in increased mean magnitudes of SCRs and affect levels of generalisation. The interference provided by the novel probe reduced levels of generalisation and negated a number of previously identified correlations between the trait questionnaires and the dependent measures, when results were directly compared to those from Experiment 3. However, Experiment 5 highlighted that there existed a clearly distinguishable cohort of participants who showed robust and reliable generalisation across all of the dependent measures despite any interference. Experiment 6 sought to discriminate between ‘generalisers’ and ‘non-generalisers’ by adding additional semantic generalisation cues (i.e., antonyms) during generalisation testing and further examine the interfering effect of additional probe stimuli. It was hoped that this group of persistent generalisers would be more likely to be discriminable from the non-generalisers using the questionnaire. Despite significant differences in the avoidance responses and generalising behaviour of both groups, a comparison of trait scores across the two cohorts revealed no significant differences for any of the trait questionnaires examined. The overall conclusion of this program of research was that while both the semantic and symbolic generalisation phenomenon have been consistently supported, correlations between anxiety, personality or experiential trait measures and the observed behaviour have resisted identification. From the evidence outlined herein, it is clear that while more and less avoidant cohorts of participants exist, they do not appear to be easily identifiable based on trait test scores.