Abstract Plasma spraying is a very complex process, controlled by a large number of process parameters. The spray gun parameters control the plasma plume and thereby the velocity and temperature of the particles in the plasma. Some of the spray gun parameters are difficult or impossible to control, but variations of them give rise to fluctuations in the microstructure of the sprayed thermal barrier coating and thereby low reproducibility. By movement of the control from the spray gun to direct control of the particle properties in the plasma this problem will be avoided, and it should result in better process control, higher quality of the final coating and thus improved reproducibility. In this study, the influence of the plasma spray process on the coating microstructure was investigated. An orthogonal factorial designed experiment was performed, where eight process parameters were varied, resulting in 16 different coatings. The particle properties were observed in-situ with the optical measurement system DPV 2000. The microstructure of the coatings was studied using optical microscopy and the amount of different features, i.e. cracks and pores, was quantified. Multiple linear regression was used to find models describing the relation between the spray gun parameters and the particle properties, between the spray gun parameters and the microstructure, and between the particle properties and the microstructure. The results showed that the spray gun parameters well describe the variation in particle velocity and particle temperature. Further, it was found that particle velocity, particle temperature, spray angle, and substrate temperature are the most important parameters concerning influence on the coating microstructure. However, their influence on the different microstructure features varied. The study implies that focus can be set on one or two particle properties measured in the plasma, instead of the numerous spray gun parameters.