Dairy farming is one of the most energy- and emission-intensive industrial sectors, and offers noteworthy opportunities for displacing conventional fossil-fuel consumption both in terms of cost saving and decarbonisation. In this paper, a solar-combined heat and power (S–CHP) system is proposed for dairy-farm applications based on spectral-splitting parabolic-trough hybrid photovoltaic-thermal (PVT) collectors, which is capable of providing simultaneous electricity, steam and hot water for processing milk products. A transient numerical model is developed and validated against experimental data to predict the dynamic thermal and electrical characteristics and to assess the thermoeconomic performance of the S–CHP system. A dairy farm in Bari (Italy), with annual thermal and electrical demands of 6000 MWh and 3500 MWh respectively, is considered as a case study for assessing the energetic and economic potential of the proposed S–CHP system. Hourly simulations are performed over a year using real-time local weather and measured demand-data inputs. The results show that the optical characteristic of the spectrum splitter has a significant influence on the system''s thermoeconomic performance. This is therefore optimised to reflect the solar region between 550 nm and 1000 nm to PV cells for electricity generation and (low-temperature) hot-water production, while directing the rest to solar receivers for (higher-temperature) steam generation. Based on a 10000-m2 installed area, it is found that 52% of the demand for steam generation and 40% of the hot water demand can be satisfied by the PVT S–CHP system, along with a net electrical output amounting to 14% of the farm''s demand. Economic analyses show that the proposed system is economically viable if the investment cost of the spectrum splitter is lower than 75% of the cost of the parabolic trough concentrator (i.e., <1950 €/m2 spectrum splitter) in this application. The influence of utility prices on the system''s economics is also analysed and it is found to be significant. An environmental assessment shows that the system has excellent decarbonisation potential (890 tCO2/year) relative to conventional solutions. Further research efforts should be directed towards the spectrum splitter, and in particular on achieving reductions to the cost of this component, as this leads directly to an increased financial competitiveness of the proposed system.