Abstract: With the widespread construction of safety monitoring facilities for small reservoir dams, how to utilize the data value of these facilities and scientifically assess the service performance of dams has become an important research topic. The seepage pressure of dams, as an important physical quantity for evaluating the seepage characteristics of earth-rock dams, is a necessary monitoring parameter for the construction of new safety monitoring facilities for small reservoirs. Accurately analyzing the sources of seepage pressure errors and improving the accuracy of data observation are the basis for achieving the scientific application of monitoring data and supporting dam safety evaluation. Previous analyses of the sources of seepage pressure errors for dams have included precipitation, water level in front of the dam, air pressure, sensor calibration parameters, and engineering construction. However, as a small-scale climate carrier, the reservoir is also affected by various meteorological factors within the region. Factors such as temperature, relative humidity, wind direction, wind speed, radiation, rainfall, and soil moisture constitute the regional microclimate, which is closely related to variations in dam seepage pressure. There are relatively few studies on changes in seepage pressure values caused by multiple meteorological factors. In order to investigate the influence of regional microclimate on seepage pressure of small reservoir dams, a regional microclimate observation system and a dam seepage pressure observation system were established, and field tests were conducted at the reservoir. These systems enabled the automatic collection of meteorological data, including soil moisture, air pressure, temperature, and relative humidity, as well as reservoir seepage pressure data. Based on monthly cumulative rainfall, data with monthly cumulative rainfall below 30 mm were selected as samples under drought conditions, while data with monthly cumulative rainfall above 100 mm were selected as samples under heavy rainfall conditions. A multiple meteorological factor partial least squares regression (PLSR) model was established, and the optimal number of components for the regression model was determined through 5-fold cross-validation. The reliability of the model was evaluated using the correlation coefficient R² and the root mean square error RMSE. Through variable projection importance (VIP) analysis, the score values of each meteorological factor were calculated, quantifying the contribution of each meteorological factor to seepage pressure. Based on the measured seepage monitoring data from two monitoring points, UP5 and UP6, on the maximum cross-section of a reservoir dam, the applicability of this method was verified. The results show that this method is reliable. Under drought conditions, the R² and RMSE values of UP5 and UP6 are 0.873 and 0.942, and 0.3536 and 0.2399, respectively. By calculating the VIP score, the VIP score of air pressure is greater than 1, indicating that it is the main factor affecting the variation in seepage pressure. Under heavy rainfall conditions, the R² and RMSE values of UP5 and UP6 are 0.832 and 0.418, and 0.4044 and 0.7524, respectively. By calculating the VIP score, the VIP scores of air pressure and soil moisture are greater than 1, indicating that they are the main factors affecting the variation in seepage pressure. By calculating the standardized regression coefficient of soil moisture under heavy rainfall conditions and combining it with the spatial and temporal distribution of monitoring data, it was found that soil moisture has a lag effect on changes in seepage pressure. Compared with existing research results, this study further reveals the influence of meteorological factors on seepage pressure variations. The research results can provide theoretical and data support for the scientific correction of seepage pressure values in small reservoir dams.