Previous research has provided evidence linking individuals' psychological traits to their ability to perceive, interpret and respond to facial expressions of emotion (i.e. the emotion decoding process). These 'trait-decoding links' have been established by testing the relationships between participants' personality and emotional intelligence (EI), and their performance in emotional expression-judgment/labelling tasks. However, recent research challenging the core concepts of these methods calls into question the ecological validity of established trait-decoding links. The present research consisted of four studies, each representing an iterative attempt to address the aim of this thesis: re-evaluating the reliability of previously established trait-decoding ability relationships, with decoding tasks that use dynamic, spontaneously elicited emotional-expressions as stimuli. Across each study, an emotion-tracing task was used to measure participants' continuous ratings of spontaneously elicited emotional-expressions (focusing on expression intensity), portrayed in a series of short video-clips. The analyses focused on establishing how the reliability, average intensity and variation of these ratings related to psychological traits, and other measures of decoding ability. Initially, the results provided tentative evidence for the validity of the emotion-tracing task as a measure of decoding ability (Study 1). However, a subsequent inability to replicate these results using an improved and expanded methodology (Study 2), prompted an investigation exploring why more conventional measures of decoding ability appeared to more reliably replicate trait-decoding links, compared to the trace-based · methodologies (Studies 3 and 4). The results indicated that the likelihood of finding trait-decoding relationships may be higher when using posed expressions of emotion as experimental stimuli. It is argued that the spontaneously elicited expressions used in the emotion-tracing tasks were less likely to engage the attention-based processes that underpin trait-decoding links. These findings suggest previous research has overestimated the validity of psychological traits as predictors of how effectively individuals can decode human emotion in day-to-day interactions.