It is proposed by scholars of consumer complaining behaviour that dissatisfaction occurs at two levels: (1) the primary level resulting from product or service failure; and (2) the secondary level resulting from an unsuccessful attempt at complaint resolution with the company held responsible. This thesis provides an integrated analysis of the triggers of voiced and non-voiced consumer responses to the level of primary dissatisfaction. A deductively derived two-factor taxonomy of consumer responses to dissatisfaction was developed out of the extant literature, to identify those responses to dissatisfaction analogous with both primary and secondary dissatisfaction. Based on this taxonomy, a deductively derived typology of consumer responses to primary dissatisfaction, together with two truncated alternatives, was also proposed. The proposed typologies facilitated two levels of analysis: (1) the unidimensional level relating to whether or not the dissatisfaction was voiced to the company held responsible; and (2) the multi-dimensional level of response style engaged in. At each level of analysis, the contribution to total explained variance of six trigger sets was assessed: pre-dissatisfaction situation, post-dissatisfaction situation, company/consumer relationship, marketplace/consumer relationship, psychographics and demographics. The validity of both the proposed two-factor taxonomy and the three typologies was examined through a cross-sectional survey of 1000 dissatisfied consumers across eight product and service categories. Findings supported the deductive basis of the taxonomy, where it was found that whilst exit behaviours and private negative word-of-mouth are analogous with the level of primary dissatisfaction, the incidence of public negative word-of-mouth, third party action and grudge holding was significantly greater among consumers experiencing secondary dissatisfaction. These findings also supported the deductive basis of the proposed typologies of responses to primary dissatisfaction, by empirically identifying those responses analogous with this level of dissatisfaction. At the unidimensional level of analysis, the trigger sets explained 54% ofthe variance between a voiced and a non-voiced instance of primary dissatisfaction. At the multidimensional level of the response style, 63% of variance was explained between response styles in the truncated typology with the most practical use for suppliers. On the basis of these findings, recommendations are made to practitioners on how to encourage supplier-friendly consumer response styles following primary dissatisfaction, and to discourage less friendly styles. The empirically supported truncated typology -distinguishing between passive, private responses, telling, and telling + private responses -represents an important strategic tool for suppliers to monitor and benchmark responses to primary dissatisfaction among their customer base.