The species-area relationship (SAR) is a core component of community ecology, and is an important basis for biological diversity scaling. The SAR is used to describe community types and can solve many ecological problems, such as the determination of minimum sampling areas in a community. Therefore, it is of great importance to diversity conservation. Recently, a number of studies have demonstrated substantial uncertainties in selecting the best SAR model for a data set. In the present study, a 30-hm² permanent forest plot was established in a broad-leaved Korean pine forest in Jiaohe, Jilin Province, China. All trees with diameters at breast height (DBH) ≥ 1 cm were tagged and the height, DBH, and crown diameter of these trees were measured and recorded. We established a logarithmic model, a power function model, and a logistic model using the 30-hm² sample plot to simulate the SAR of a broad-leaved Korean pine forest. We examined how SARs simulated by logarithmic, power function, and logistic models differed after random sampling or nested sampling methods had been used to collect data,and how this difference was affected by sampling scales (broad, moderate, and fine scales). The Akaike Information Criterion (AIC) value was used to compare the goodness-of-fit for each SAR model. The results showed that the sampling method had a significant influence on the SAR, and that the goodness-of-fit for random sampling was better than that for nested sampling at all sampling scales. The establishment of a species-area relationship was closely related to the sampling scales, and the logarithm and logistic models were superior to the power function model at the fine scale (10 hm²). At the moderate and large scales (20 hm² and 30 hm², respectively), the logistic model better fitted the species-area relationship for broad-leaved Korean pine forest than did the logarithm and the power function models..A comparison of the different models showed that the logistic model with random sampling produced an optimal fit for the species-area relationship within the 30 hm² broad-leaved Korean pine sample area (AIC =76.91), and that the appropriate minimum sampling area was 10 hm². We concluded that both sampling scale and sampling method had significant influences on the SAR. The scale effect on the SAR is closely related to the community species distribution pattern, and the impacts may result from habitat heterogeneity and successional stage. Habitat heterogeneity and community succession stage might have influenced the number of regional species and species composition, and these different species distribution patterns were reflected in the different SAR curves. Therefore, in practical applications, the variation in the actual community structure and environments within the sampling area should be fully considered. Further work needs to consider the actual situation of the local forest community to simulate the species-area relationship models