The rates of transmission of infection from inoculum or infecteds to susceptible hosts are critical determinants of epidemics, yet no formal experimental methods have been described for their quantification and analysis in spatially explicit epidemics. Replicated microcosms of >400 radish seedlings and with tight control of environmental conditions were exposed to known amounts of inoculum of the fungal plant pathogen Rhizoctonia solani. Spatiotemporal maps of disease progress were used to distinguish between primary and secondary infections and to count changes with time in the number of infected plants and the number of contacts between susceptible and neighboring infected plants. Transmission rates were defined within a compartmental S–I (susceptible–infected) model for plant epidemics and estimated empirically using counts from spatial maps. The transmission rate for primary infection declined with time; the transmission rate for secondary infection rose initially and then declined. We discuss the mechanisms contributing to the changes in transmission rates with time and show that spatial mapping combined with an epidemiological analysis provides accurate empirical estimates of transmission rates.