The invention relates to a terahertz imaging-based biomacromolecule content relative quantitative detection method, and belongs to the field of biomacromolecule detection. The method comprises the following steps: S1, obtaining a nitrocellulose film or a polyvinylidene fluoride film containing biomacromolecules; S2, placing the to-be-detected film on a terahertz time-domain spectrometer, and carrying out scanning imaging to obtain a time-domain spectrum of a point; S3, performing Fourier transform and smoothing processing on the time domain spectral data, and taking the processed time domain spectral data as frequency domain spectral data; S4, taking N points on each to-be-detected film, measuring the thickness of the to-be-detected film, and taking a mean value as the thickness of the to-be-detected film; and S5, calculating, analyzing and recombining the frequency domain spectral data of each pixel point of the measurement film, and converting the frequency domain spectral data intogray values of the corresponding pixel points, so as to form a visual image reflecting the spatial distribution condition of the biomacromolecule content on the film. The method has the advantages ofno need of antibodies, simple operation, fast and accurate detection speed, time saving, labor cost saving, and realization of label-free detection.