As a result of a court order, computerized summaries of approximately 1,300 cases reported as Guillain-Barré syndrome by state health departments to the Centers for Disease Control during the intensive national surveillance instituted following the swine influenza vaccination program in 1976-1977 became available for further study. Although the data were not uniformly adequate to confirm the diagnosis of Guillain-Barré syndrome, they were sufficient to enable classification according to extent of motor involvement. Vaccinated cases with "extensive" paresis or paralysis occurred in a characteristic epidemiologic pattern closely approximated by a lognormal curve, suggesting a causal relationship between the disease and the vaccine. Cases with "limited" motor involvement showed no such pattern, suggesting that this group included a substantial proportion of cases which were unrelated to the vaccine. The effect attributed to the vaccine lasted for at least six weeks and possibly for eight weeks but not longer. The relative risk of acquiring "extensive" disease over a six-week period following vaccination ranged from 3.96 to 7.75 depending on the particular baseline estimate of expected normal or endemic incidence that was chosen. Correspondingly, the number of cases that could be attributed to the vaccine over the six-week period ranged from 211 to 246, or very slightly higher over an eight-week period if the lowest baseline estimate was used. The total rate of Guillain-Barré syndrome cases attributed to prior use of the vaccine was 4.9 to 5.9 per million vaccinees.