Therapeutic injection of facet joints is now widely practised, but British experience has been infrequently reported. We studied the results of injecting facet joints with a corticosteroid preparation in 50 patients suffering from the "facet syndrome". Our series included a number of extra-articular injections and these "failed injections" provide a useful control group. Results indicate that only intra-articular injections are effective; certainty of joint penetration can be ensured only by the routine use of joint arthrography.