Context. The Perseus OB1 association, including the h and χ Persei double cluster, is an interesting laboratory for the investigation of massive star evolution as it hosts one of the most populous groupings of blue and red supergiants (Sgs) in the Galaxy at a moderate distance and extinction.
Aims: We discuss whether the massive O-type, and blue and red Sg stars located in the Per OB1 region are members of the same population, and examine their binary and runaway status.
Methods: We gathered a total of 405 high-resolution spectra for 88 suitable candidates around 4.5 deg from the center of the association, and compiled astrometric information from Gaia DR2 for all of them. This was used to investigate membership and identify runaway stars. By obtaining high-precision radial velocity (RV) estimates for all available spectra, we investigated the RV distribution of the global sample (as well as different subsamples) and identified spectroscopic binaries (SBs).
Results: Most of the investigated stars belong to a physically linked population located at d = 2.5 ± 0.4 kpc. We identify 79 confirmed or likely members, and 5 member candidates. No important differences are detected in the distribution of parallaxes when stars in h and χ Persei or the full sample are considered. In contrast, most O-type stars seem to be part of a differentiated population in terms of kinematical properties. In particular, the percentage of runaways among them (45%) is considerable higher than for the more evolved targets (which is lower than ∼5% in all cases). A similar tendency is also found for the percentage of clearly detected SBs, which already decreases from 15% to 10% when the O star and B Sg samples are compared, respectively, and practically vanishes in the cooler Sgs. Concerning this latter result, our study illustrates the importance of taking the effect of the ubiquitous presence of intrinsic variability in the blue-to-red Sg domain into account to avoid the spurious identification of pulsating stars as SBs.
Conclusions: All but 4 stars in our working sample (including 10 O giants/Sgs, 36 B Sgs, 9 B giants, 11 A/F Sgs, and 18 red Sgs) can be considered as part of the same (interrelated) population. However, any further attempt to describe the empirical properties of this sample of massive stars in an evolutionary context must take into account that an important fraction of the O stars is or likely has been part of a binary/multiple system. In addition, some of the other more evolved targets may have also been affected by binary evolution. In this line of argument, it is also interesting to note that the percentage of spectroscopic binaries within the evolved population of massive stars in Per OB1 is lower by a factor 4-5 than in the case of dedicated surveys of O-type stars in other environments that include a much younger population of massive stars.

Tables A.1-A.5 are only available at the CDS via anonymous ftp to http://cdsarc.u-strasbg.fr (ftp://130.79.128.5) or via http://cdsarc.u-strasbg.fr/viz-bin/cat/J/A+A/643/A116