The objective of this paper is to develop non-linear proportional-integral-plus (PIP) control algorithms for regulating ventilation rate in mechanically ventilated agricultural buildings. State-dependent parameter (SDP) models are developed for an environmental test chamber, representing a section of a livestock building or glasshouse. Here, the system is modelled using the quasi-linear SDP model structure in which the parameters are functionally dependent on other variables in the system. The model is subsequently utilised to develop a new approach to control system design, based on non-linear PIP pole assignment, with a discrete-time Smith Predictor to handle the sampled time delays. Implementation results for the test chamber demonstrate improved control of ventilation rate, with a faster response to disturbances in comparison with both linear and conventional (linearised) scheduled PIP control. The approach has application to a wide class of other non-linear systems, as demonstrated by simulation examples.