The problem of forecasting the dynamics of the development of solar energy in the region in the context of the global trend of energy transition is considered. The urgency of the problem is due to the fact that forecasts of the development of solar energy are usually characterized by relatively large errors. To solve this problem, the author proposed a multi-trend approach to constructing a regional function for the growth of solar power capacity. The method is based on the description of the dynamics of power growth in the form of the average value of logistic, linear and exponential trends. The weighting factors are equal to values that are inversely proportional to the errors of the corresponding trends. Based on this method, forecasts of solar energy capacity were calculated for Africa, Asia, Europe, North America and South America for the period 2017–2019. The validity of the method is confirmed by the fact that these forecasts are characterized by a relatively low deviation from the actual data. The author has developed a forecast for these regions for the period 2020-2023. It is shown that the reason for the low reliability of most forecasts is the desire to use the logistic curve as a universal analysis tool. This approach absolutes the logistics trend and does not take into account the specifics of the region. However, for some regions, a linear or exponential trend can serve as the dominant growth trend in solar energy capacity. In particular, the reason for the systematic underestimation of forecasts for China was the ignorance of the exponential component of the growth of solar energy capacity.