In the two studies, "self-harm" (SH) is defined as "any form of self-injurious behavior, regardless of intention to die or not" (Fortune and Hawton, 2005). This definition recognizes that motivations can often be mixed or ambivalent, and is widely used in European studies. In one study, logistic regression analysis found that explicit (thoughtful, self-reported) rather than implicit measures discriminated between SH cases and controls, which differs from the findings of previous research. Faster reaction times were observed for negative implicit associations (SH and "I dislike") compared to positive implicit associations (SH and "I like") for both SH cases and controls. The SH group was faster to respond to arousal implicit associations compared to implicit sedation association. The second study extended this finding in order to demonstrate associations between evaluative/arousal Go No-Go associations and self-reported functions of SH. Internal motivations for SH behavior were significantly related to an implicit sedation association with SH; whereas, interpersonal motivations were associated with an implicit arousal association with SH. These findings support existing functional accounts of SH and provide novel experimental support for the hypothesized role of automatic emotion regulation and social/interpersonal functions of SH. These findings suggest that interventions that consider only the deliberative level of processing may be ineffective, as this may not address the problems at the associative level. Determining whether or not individuals are consciously aware of such automatic thoughts and whether they are able to suppress them is necessary in efforts to reduce SH. Study limitations are discussed. 6 tables and 45 references