SUMMARYExperiments involving direct placement of inoculum at various distances from an infection court are used to estimate the probability of infection and/or disease of seminal roots of wheat by Gaeumannomyces graminis (Sacc.) Arx & Olivier var. tritici Walker and of germinating seeds of radish by Rhizoctonia solani Kuhn. Two concepts underlie the relationships under investigation. These are the pathozone, which is characterized by the furthest extent of inoculum from the host from which infection is recorded {z), and infection efficiency, which is characterized by two parameters a and /?. The asymptotic infection efficiency, /?, is estimated by the proportion of propagules that cause infection when placed in contact with the host; a is a measure of the decline in infection efficiency of propagules of inoculum with increasing distance from the host. Three non-linear models are used to describe the profile of infection efficiency over distance.The sensitivity of the method of placement of inoculum in detecting treatment effects is demonstrated first by showing differences in response amongst two isolates of G. graminis and two times of observation. Maximal values of z for G. graminis ranged from 4 to 8 mm, depending upon the type of inoculum of G. graminis and the growth medium used. Values of z for R. solani and germinating radish seeds varied from 12 to 16 mm for propagules comprising fragments of chopped potato soil inoculum sieved to 1 to 2 mm.The amounts and nutritional basis of inoculum infiuenced the probability of infection and the parameters a, y? and z. Thus, for G. graminis, estimated values for oT^, /? and z were greater for inoculum comprising colonized millet grains than for colonized sand particles. When seedlings of radish were exposed to clumps of propagules of R. solani, the parameters /? and z increased, while a decreased with increasing clumip density. Reduction in propagule size to 05 to 10 mm reduced the proportion of infected bosts. When clumps of seedlings of radish were challenged with clumps of inoculum of R. solani, the probability of infection of clumps increased with clump density, but there was no evidence for an effect on the probability of infection of individual seedlings within a clump. The relationship between the estimates of pathozone width and infection efficiency and models for infection of subterranean organs is discussed.