The pathogenicity of Erwinia carotovora pv. carotovora, E. carotovora pv. atroseptica and E. chrysanthemi to potato plants was examined in experimental and commercial crops in Scotland and Israel in 1983 and 1984. Symptoms caused by the three erwinias tended to be different depending on when they were expressed during the different growing seasons. Field symptoms were grouped into four types based on the origin and extent of stem rot and of leaf desiccation. Most were reproduced on inoculated potted plants held under different combinations of relative humidities (ca 70 and ca 100%) and temperatures (17° and 30°C). In plots planted with seed inoculated with E. c. atroseptica alone or together with E. c. carotovora or E. chrysanthemi (in Israel only), stem rot symptoms developed throughout the growing season in Scotland (Invergowrie) but only before May in February-sown crops and after November in September-sown crops in Israel (Negev). In contrast, E. chrysanthemi alone or together with one or the other erwinias in Israel caused extensive leaf desiccation with no stem rot only in May and before December in the spring-and autumn-sown growing seasons, respectively. None of 13 E. c. carotovora strains inoculated into erwinia-free tubers was pathogenic in the spring growing season in Israel. More than one erwinia type was frequently present in both healthy and diseased stems, usually the same bacteria which were inoculated in ihe seed tubers. Isolations made from diseased stems showed that the temporal patterns of detection frequency of E. c. atroseptica and E. chrysanthemi in stem rots and in plants with leaf desiccation symptoms were similar to those for expression of the two symptom types. Erwinia carotovora pv. carotovora alone was never found in diseased stems although it could be isolated together with the other erwinias in both Scotland and Israel. Similar results were obtained when diseased stems from commercial potato fields were examined, except that E. c. carotovora was the sole erwinia present in aerial stem rots. The patterns observed in the pathogenicity of three erwinias were tentatively explained in terms of temperature changes during the growing season, the transition points being in the region of 20°C for maximum soil temperature at tuber depth or 25°C for maximum air temperature above which E. c. atroseptica and below which E. chrysanthemi were not apparently pathogenic.