Spectral analysis as a powerful mean to identify the characteristics of data series is introduced in this paper. The problems requiring further explorations in grey system theory are also identified. This includes discrimination of various factors of a data sequence in frequency domain, spectral analysis of various sequence operators, the synthesis axiom of degree of greyness for “multiplication” and “division” etc. Further, how to select and test a grey prediction model? How to select and test an inverse grey incidence analysis model? The test rules and criteria of grey clustering evaluation models, etc.