I have carried out a search for Planet 9 in the IRAS data. At the distance range pro-posed for Planet 9, the signature would be a 60 µm unidentified IRAS point source with an associated nearby source from the IRAS Reject File of sources which received only a single hours-confirmed (HCON) detection. The confirmed source should be detected on the first two HCON passes, but not on the third, while the single HCON should be detected only on the third HCON. I have examined the unidentified sources in three IRAS 60µm catalogues: some can be identified with 2MASS galaxies, Galactic sources or as cirrus. The remaining unidentified sources have been examined with the IRSA Scanpi tool to check for the signature missing HCONs, and for association with IRAS Reject File single HCONs. No matches of interest survive. For a lower mass planet (65ME) in the distance range 200-400 AU, we expect a pair or triplet of single HCONs with separations 2-35 arcmin. Several hundred candidate associations are found and have been examined with Scanpi. A single candidate for Planet 9 survives which satisfies the requirements for detected and non-detected HCON passes. A fitted orbit suggest a distance of 225±15 AU and a mass of 3-5 ME. Dynamical simulations are needed to explore whether the candidate is consistent with existing planet ephemerides. If so, a search in an annulus of radius 2.5−4ocentred on the 1983 position at visible and near infrared wavelengths would be worthwhile.