Spleen tyrosine kinase (Syk) is a cytoplasmic tyrosine kinase involved in signalling in many of the cells that drive immune inflammation. The development of small molecules that inhibit Syk kinase may change the way we treat disorders such as rheumatoid arthritis (RA), as well as a range of other inflammatory diseases. Fostamatinib (R-788) is an orally bioavailable small molecule. It is the prodrug of R406, which is a potent Syk inhibitor. Fostamatinib was developed because it has more favourable physiochemical properties. It is rapidly converted to R406 by intestinal enterocytes. It has been evaluated in experimental models of RA, such as collagen-induced arthritis. In these models, fostamatinib suppressed clinical arthritis, bone erosions, pannus formation and synovitis. A phase II programme with fostamatinib has largely been completed. Three key trials have been published, lasting 12-26 weeks and each enrolling 189-457 patients (875 in total). All these trials involved placebo therapy and patients continued to receive methotrexate in addition to active treatment with fostamatinib. The first dose-ranging trial evaluated three treatment doses in RA patients who had not fully responded to methotrexate therapy. The second trial compared two treatment doses in patients who had not responded to methotrexate therapy. The third trial compared a single treatment dose with placebo in patients who had not responded to biological therapy. The primary outcome measure was the number of patients achieving American College of Rheumatology (ACR) 20% (ACR20) responses. Placebo ACR20 response rates in all three trials were similar (35-38%). All three trials involved one treatment arm receiving fostamatinib 100 mg twice daily; ACR20 responses with this active treatment ranged from 38% to 67%. A meta-analysis of ACR responses in these trials, using responses to the highest dose in each trial for comparisons with placebo therapy in a random effects model, showed a borderline benefit with ACR20 responses. There were more significant differences with ACR50 and ACR70 responses. The reason that this meta-analysis was not more strongly positive is that the third trial, which evaluated patients who had failed to respond to biological treatments, gave negative results. Individual ACR response components, such as changes in swollen joint counts, showed significant differences in the first two trials, but there were no definite treatment benefits in the third trial. Overall, the differences were significant in a meta-analysis of all three trials. The most important adverse reactions were diarrhoea, neutropenia and raised ALT levels, which all showed significant excesses with active treatment compared with placebo. Too few patients have been studied for a definitive safety profile to be known. Overall, the results of the phase II trials were sufficiently encouraging for a phase III programme to be initiated. It will be some years before their definitive results are available.