The cognitive processing of subjective intensity information in the judgement of subjective similarity between pairs of complex odour mixtures is discussed in relation to studies of odour intensity, odour pleasantness, subjective similarity in general, and olfactory classificatory schemes in particular. Preliminary experiments dealt with olfactory intensity per se, focussing on scaling method, response variability, masking phenomena, and led to the construction of alternative simulation models for combining intensity information to generate theoretical similarity data. Traditional practices for scaling subjective intensity are questioned by these data. A pilot study of subjective similarity revealed consistent individual differences in judgement criteria and multidimensional scalings and permissible interpretations thereof. A second similarity experiment disclosed the confounding of context effects in perceived similarity, revealed the role of intensity in qualitative similarity, and questioned the usefulness of perceived similarities as a basis for classification of odour quality.