AbstractInjection of wind power into an electric grid affects the voltage quality. As the voltage quality must be within certain limits to comply with utility requirements, the effect should be assessed prior to installation. To assess the effect, knowledge about the electrical characteristics of the wind turbines is needed or else the result could easily be an inappropriate design of the grid connection. The electrical characteristics of wind turbines are manufacturer‐specific but not site‐specific. This means that, having the actual parameter values for a specific wind turbine, the expected impact of the wind turbine type on voltage quality when deployed at a specific site, possibly as a group of wind turbines, can be calculated. The methodology for this is explained and illustrated by case studies considering a 5 × 750 kW wind farm on a 22 kV distribution feeder. The detailed analysis suggests that the wind farm capacity can be operated at the grid without causing unacceptable voltage quality. For comparison, a simplified design criterion is considered assuming that the wind farm is only allowed to cause a voltage increment of 1%. According to this criterion, only a very limited wind power capacity would be allowed. Measurements confirm, however, the suggestion of the detailed analysis, and it is concluded that a simplified design criterion such as the ‘1% rule’ should not be used for dimensioning the grid connection of wind farms. Rather, this article suggests a systematic approach including assessment of slow voltage variations, flicker, voltage dips and harmonics, possibly supported by more detailed analyses, e.g. system stability if the wind farm is large or the grid is very weak, and impact on grid frequency in systems where wind power covers a high fraction of the load, i.e. most relevant for isolated systems. Copyright © 2002 John Wiley & Sons, Ltd.