Estimates of the size of the EU cannabis market vary widely, with figures ranging from €15 billion to €35 billion per year. An important source of uncertainty is the limited information available about typical quantities consumed by different types of users. Indeed, some studies focused on the EU or specific Member States rely on studies of cannabis users outside of the EU. This report generates estimates of retail cannabis expenditure in the EU using new data about cannabis consumption and expenditures from a web survey conducted in seven Member States: Belgium, the Czech Republic, Italy, Netherlands, Portugal, Sweden, and the United Kingdom (n=4,156). Combining insights from these surveys with data collected by the EMCDDA about cannabis prevalence, frequency, and prices, this report suggests the market may be considerably smaller than previously estimated. Calculations here suggest a range of approximately €7 billion to €10 billion annually circa 2010 before accounting for the "consumption gap" that is created when data from general population surveys are used to measure substance use. Evidence from the alcohol literature suggests that true consumption can be as much as double that estimated from general population surveys. Thus, these estimates are likely low. Information about the type of cannabis consumed (herbal versus resin) across countries is scant, but if the new estimates published by the EMDCCA (2012) are correct, they suggest that roughly 50-65% of all cannabis consumed in the EU is resin. In addition to generating figures that should be of interest to policymakers, the analyses presented here make important methodological contributions. For example, we demonstrate that since consumption intensity (grams per day of use) is positively correlated with consumption frequency (days used per month), multiplying the average number of use days by the average number of grams consumed per use day generates consumption figures that are lower than the correct approach of multiplying each individual's days consumed and daily consumption figures and then averaging across individuals only after that multiplication.