Abstract The choice of a suitable theoretical wind speed distribution is an important prerequisite for accurate wind energy yield estimation. In this review, 46 studies published 2010–2018, which compared the goodness-of-fit of different theoretical parametric distributions, were evaluated. The evaluation scheme considered three aspects: (1) distributions, (2) parameter estimations methods, and (3) goodness-of-fit metrics. It was found that the two-parameter Weibull distribution is by far the most frequently (in 44 out of 46 studies) evaluated distribution. In total, 115 different distributions were fitted to wind speed data. Out of these 115 distributions, 32 distributions were recommended at least once. One reason for the large spread of recommendations can be attributed to the research design. To ensure comparability in this review, each study was rated by its scope and the amount and quality of the wind speed data used. Results from the present evaluation demonstrate that the five-parameter Wakeby and four-parameter Kappa distributions achieved the highest scores. The results also show that the maximum likelihood method, the least-squares estimation method, the moment method, and the L-moment method are the most frequently applied methods for estimating distribution parameters. For goodness-of-fit evaluation, many studies used more than two metrics. Among the most often used goodness-of-fit metrics are the Kolmogorov-Smirnov statistic, the coefficient of determination, and the Chi2 statistic. A fundamental conclusion of this review is that common research standards are needed to improve the comparability of future wind energy yield assessments. Standards are required for wind speed data, the use of distributions, and the goodness-of-fit evaluation.