Attempts to account for the variations in photosystem II (PSII) under general conditions result in non-linear and cumbersome models that are difficult to validate and render few insights about the system kinetics. In this research, the authors experimentally show that under certain conditions, linear-system techniques could be applied to advantage for probing some basic kinetic characteristics of the plastoquinones (PQs). The PQ redox states of the reaction centres were represented in a conditionally linear model structure with delayed fluorescence (DF) as a measurable output. DF data were acquired for different plant samples and conditions. After least-squares parameter optimisation, not only could the model closely describe the measured DF, but more significantly, the estimated parameters correctly reflected the expected changes induced by drought or [3-(3,4-dichlorophenyl)-1,1-dimethylurea] (DCMU) stress. Analysis showed that for short-pulse illumination, the PQ kinetic states of the reaction centres in an initially dark-adapted plant leaf can be represented as a time-invariant bilinear system in a five-dimensional state space. The system becomes linear for constant illuminations, but the system matrix and the kinetic behaviour are illumination dependent. In particular, the system behaves differently between lights-on and lights-off conditions. The simplicity of the model structure, nonetheless, permits observation and analysis of the PQ kinetics of PSII reaction centres from DF measurements by using linear-system techniques.