31 Kreibig, S. D. Brosch, T. Schaefer, G. 9998 K. R. Scherer T. Baenziger E. Roesch A blueprint for an affectively competent agent: Cross-fertilization between emotion psychology, affective neuroscience, and affective computing Oxford, England Oxford University Press xxx SCAS, affective computing, psychophysiology In this chapter, we introduce concepts relevant to emotion measurement in the domain of autonomic nervous system (ANS) activity. In the first part, we review theoretical accounts of why and how emotions are believed to influence ANS activity and what we can learn about emotion from monitoring the ANS. The notion of indicand (an abstract quality), indicator (a concrete measure of that quality), and the nature of their relation are introduced for conceptualizing the intersecting field of affective computing and autonomic physiology. In the second part, we turn to the question of which ANS parameters to measure. Based on the structure of the ANS, measures that index activation of the cardiovascular, respiratory, and electrodermal organ systems are introduced and sensor concepts are reviewed. In the third part, we address the question of how to design a particular measurement. Practical issues of the differentiation between emotion detection and emotion identification, characteristics of the measurement context for collecting physiological data, as well as aspects of physiological data analysis are addressed. We also give consideration to possible application contexts afforded by the measurement of emotion via ANS activity as well as ethical considerations in the physiological measurement of emotion. In conclusion, we offer a set of 'golden rules' of physiological measurement for affective computing as a practical guide to researchers and engineers.