This publication is one of two thematic journal issues produced by the project Trouble per liter - The effects of policies favoring lighter beverages. The project's origins date back to March 2007, when Nordic and Russian alcohol researchers met up for a research-based alcohol policy conference in Moscow. During the meeting it became evident that, from a Russian point of view, the historic Nordic drinking trend away from a domination of distilled spirits towards a more mixed pattern, with beer playing a major role, could be interpreted as a successful result of a long-term alcohol policy strategy. Had a conscious steering of the overall consumption towards a preference of lighter-alcohol-content products indeed led to a more moderate and less harmful consumption style in the Nordic countries? The Nordic researchers were struck by the lack of evidence in this area. Could the claim of a successful policy be proven true or false - had larger shares of lighter beverages decreased overall rates of alcohol-related harm - even if the effect was measured, for example, per liter of ethanol consumed?To look into the matter, an international project group was established in 2009 with the support of the Nordic committee of senior civil servants of the social and health field.* At a scientific meeting in Copenhagen in the spring of 2011, results were presented from the separate inquiries. In this issue of Contemporary Drug Problems, five studies by members of the project group examine how consumption of different beverage types are connected to harm. While Ramstedt and Boman, Kerr and Ye, Makela, and Landberg frame their inquiries in a country-specific context, Room et al. investigate study material collected from 19 countries. Each addresses the same issue, i.e. the relationship between beverage choice and harm from drinking, but each from a different viewpoint and type of data. Kerr and Ye review United States studies that have conducted time-series analyses on the connection between per capita consumption of different beverages and various harm rates, and present a new such analysis for motor vehicle accident mortality. Landberg uses time-series data from Sweden to ask whether different beverages have had a different temporal correlation with harm rates, and whether spirits drinking has had an effect on harms over and above the effect of total consumption. Makela, using both aggregate-level and individual-level data, also takes a long-term view, but instead of studying beveragespecific consumption changes, she looks at the history as a test case of whether the change towards milder beverages has brought about the wished-for moderation of drinking patterns and a lowering of harms per liter of total alcohol consumption. Ramstedt and Boman use cross-sectional individual-level data from Sweden to study the interrelationships between beverage choice, drinking patterns, characteristics of drinkers, and harm from drinking. Room et al. present a cross-cultural comparison across a number of countries of the cross-sectional associations between beverage type and harms. On the basis of the results and discussions in the articles, conclusions have been drawn in a summary by Makela et al. at the end of the issue.As a group of studies, these articles illustrate the importance of posing beverage-specific questions in alcohol research. After all, we do not say that we will go out for a glass of alcohol, but rather for a glass of wine or a beer. Likewise, when policy decisions are made regarding taxation or availability restrictions, these are often made on particular beverage types, or groups of products, rather than expressed as restrictions on alcohol.There is a long-lived dream of a better order (Olsson, 1990, p. 193) in the Nordic countries, with the goal epitomized by the wine-drinking cultures of Southern Europe. According to this vision, alcohol plays a central role in such societies, but a less dramatic role, with fewer incidents of intoxication and alcohol-related violence. …