We investigated the seasonal pattern of stroke using the Lehigh Valley Stroke Register. This register includes all patients hospitalized with stroke or transient ischemic attack (TIA) from among the 600,000 Lehigh Valley residents. Meterological data were obtained from the National Oceanic and Atmospheric Administration. The study, which uses 18 months of data, included 1,944 cases. Using single harmonic regression analysis, the seasonal pattern of TIA and infarction, but not hemorrhage, fit a sine-cosine wavefunction with a 12-month period (R2 = 41% and 36%, respectively). For infarction, the strongest seasonal pattern was exhibited for women of all ages and for both sexes in the age groups 65-74 and 75-84, but only the sine component was significant. The peak months for TIA were June-August, while the peak months for infarcts were February-April. Correlations between ambient temperature and each type of stroke were computed. A significant positive correlation for TIA was found (r = 0.57, p = 0.01). After adjusting for a 2-month lag between the low for infarction and the peak for temperature, a significant negative correlation was found (r = -0.64, p = 0.01). No significant correlation was found for hemorrhage. Possible reasons for the opposite relations of TIA and infarct are discussed.