Abstract: Seventeen isolates of B. dothidea, the cause of poplar canker, from different regions and hosts were inoculated on poplars in labs. Results statistically analysed showed the pathogens had pathogenicity differentiations. Isolates from different regions or from different hosts showed no evident differences on pathogenicity. This seemed that pathogenicity differentiation was not associated with the distribution or the host. Seventeen isolates were divided into 6 patho genie types according to their pathogenicity on five different kinds of poplars. There were some interactions between pathogenic types and hosts. Estimations of host susceptibilities to populations of B. dothidea were made. Seventeen isolates were divided into intense, middle and weak pathogenicity degrees which were expressed by numbers 1, 3, and 5 respectively. Percentages of isolates of different pathogenicity degrees in all the 17 isolates on each host were collected. These percentages of three different degrees were multiplied by the three repre-; senting numbers respectively, the products of the multiplications were added together and the sums were taken as the values for the susceptibilities of the hosts. When the value was within 1-1.4, the host was estimated to be resistant, 1.5-2.4, to be medium resistant, 2.5-3.4, medium susceptible;3.5-4.4, susceptible, 4.5-5, most susceptible. It is considered rational to use this me thod to estimate the disease resistance of hosts to pathogen populations in which pathogens have some pathogenicity differentiations.