A space-time permutation scan statistic was developed for early detection of disease outbreaks using only case data, without requiring population-at-risk information. The method adjusts for natural spatial and temporal variation and can detect outbreaks of any size and location. It was tested using emergency department visit data from New York City, where four of five strongest signals were likely local precursors to citywide outbreaks. The method was implemented in the SaTScan software and used for daily surveillance. It was found to be effective in detecting outbreaks, with minimal false signals. The method is particularly useful for areas where population-at-risk data is unavailable or unreliable. It can detect outbreaks that start locally but not those that occur simultaneously across the entire surveillance area. The method is not suitable for detecting very small outbreaks or those with mild symptoms that do not lead to emergency department visits. It is an important tool for health departments setting up early disease detection systems. The method was developed as part of New York City's surveillance initiatives and is now used daily for emergency department data analysis.A space-time permutation scan statistic was developed for early detection of disease outbreaks using only case data, without requiring population-at-risk information. The method adjusts for natural spatial and temporal variation and can detect outbreaks of any size and location. It was tested using emergency department visit data from New York City, where four of five strongest signals were likely local precursors to citywide outbreaks. The method was implemented in the SaTScan software and used for daily surveillance. It was found to be effective in detecting outbreaks, with minimal false signals. The method is particularly useful for areas where population-at-risk data is unavailable or unreliable. It can detect outbreaks that start locally but not those that occur simultaneously across the entire surveillance area. The method is not suitable for detecting very small outbreaks or those with mild symptoms that do not lead to emergency department visits. It is an important tool for health departments setting up early disease detection systems. The method was developed as part of New York City's surveillance initiatives and is now used daily for emergency department data analysis.