Emissions are an important aspect of a pellet heating system. Low harmful emissions, particularly carbon monoxide, are a measure of a well performing system. High carbon monoxide emissions are often caused by unnecessary cycling of the burner and when the average load is below the lowest possible combustion power of the burner. Combining pellet heaters with a solar heating system can significantly reduce cycling of the pellet heater and avoid the inefficient summer operation of the pellet heater. Five combined systems representing the range of typical solutions of this system type and one recently developed system have been studied, modelled and simulated. These systems are compared to a reference system, which is based on a pellet boiler and is not combined with a solar heating system. The aim was to study CO-emissions of the different types of systems and to analyse the potential of CO-emission reduction when the pellet heater is combined with a solar heating systems. Another aim was to compare the yearly CO-emissions obtained from simulations under realistic dynamic conditions with the yearly CO-emissions calculated based on the values that are obtained by the standard test methods. The study was performed with the simulation tool TRNSYS. The parameter used in the study have been identified from lab measurements on existing pellet boilers/stoves and solar heating systems. The results from the simulations show that it is possible to almost halve the CO-emission if the pellet heater is combined with a solar heating system. The results also show that the CO-emission of existing combined solar and pellet heating systems can be drastically reduced if the pellet heater is properly controlled and some basic design rules are observed. This can also be seen when analyzing the results for the new system concept where these rules have been taken into account. Comparing the yearly CO-emissions obtained from the simulations with the yearly CO-emissions calculated based on the standard test methods shows that using the latter give too low CO-values for the whole year. It is also shown that for the existing systems the average emissions under these realistic annual conditions were greater than the limit values of two Eco-labels.