Commercial samples of celery seed have been submitted to the Official Seed Testing Station, Cambridge, from 1926 to 1958, for tests to determine the percentage infection by Septoria leaf spot (Septoria apii-graveolentis and S.apii). The disease test and a method for estimating the viability of Septoria spores is described. The percentage of samples with less than 1 % infection increased from 6.1 % in 1926- 31 to 46.2 % in 1953- 8 and the percentage of heavily infected samples decreased. But celery samples submitted since 1949 for purity and germination only were more heavily infected (in 1953-8 only 26.3 % had less than 1 % infection). In 1957-8, 78.8 % of all celery seed samples tested showed evidence of infection with Septoria spp. No effect of variety or season on the incidence of disease was found during the survey. Six out of thirty samples were found to carry viable Septoria as estimated by germination of pycnidiospores. The similarity between the percentage germination of healthy seeds and those bearing Septoria pycnidia confirmed the general belief that Septoria spp. do not affect germination.