AbstractA model for OJ 287 consisting of two orbiting black holes has been constructed using optical light curve data. The model has successfully predicted the occurrence of sharp optical outbursts of OJ 287 for the past 15 years. Here we test if also the variations in the radio jet position angle can be explained within the framework of this same model, which has most of its parameters fixed by the timing of the optical flares. The model applied here has only three free parameters left, the (trivial) zero point of the jet position angle, the time lag between changes in the disk and jet orientations, and the zero point of the viewing angle. Despite its simplicity and the small number of free parameters, the model appears to be able to reproduce the main properties of the observed position angle variations during the past 30 years. The best fits are obtained when the time lag is either ~4 or ~14 years. However, the jet orientation seems to be unrelated to the direction of the spin of the primary black hole. This implies, assuming that the basic model is correct, that the mean orientation of the jet is determined by the orientation of the inner accretion disk, not by the spin axis of the black hole.