The artificial plantations is a typical terrestrial ecosystem on the island of Miaodao Archipelago, which plays a key role in regional ecological environment. The Pinus thunbergii and Pinus thunbergii×Robinia pseudoacacia are the typical trees on the Beichangshan island of Miaodao Archipelago in Shandong Province. But they are often disregarded due to their ecosystem services, particularly the carbon sequestion, which are not well understood or quantified. This paper is to estimate the trees carbon storage on the island with the biomass allometric equation and the data investigation in stand sample plot, and discuss the impact of different parameters in the island on trees carbon storage. The results showed that the average carbon storage was 84.00 t/hm² for the Pinus thunbergii, which was close to the average level in the world (86.00 t/hm²), and the average carbon storage was 29.60 t/hm² for the Pinus thunbergii×Robinia pseudoacacia, which was higher than that of the average carbon stock in Shandong Province (27.62 t/hm²). The average carbon storage of the Pinus thunbergii was higher than that of the Pinus thunbergii×Robinia pseudoacacia, which the difference was significant (P < 0.05). The allocation of carbon storage in their different organs for the Pinus thunbergii and mixed forest in tree layer was trunk> root> branch> leaf. The Pinus thunbergii forests is more suitalbe for carbon storage than that of Pinus thunbergii×Robinia pseudoacacia forests on the Beichangshan island. In order to investigate the impact of different parameters in the island on trees carbon storage, a correlation matrix including environmental factors and soil texture was calculated. The carbon storage in the island forests showed significant (P < 0.05) correlations with soil class, slope, aspect, elevation. However, all these parameters were interrrelated, and cannot be regarded as independent determining factors. Therefore, the principal component analysis (PCA) was carried out in order to extract the main factors controlling carbon storage in the island forests. The PCA extracted two factors that explained 50% of the total variance. Factor 1 was characterized by high loadings of the parameters clay soil (-0.898), silt soil (0.893), sand soil (0.922). Factor 2 was driven by slope (0.770), aspect (-0.722), elevation (0.946). A multiple linear regression model revealed factor 1 as most important factor (Beta value of 0.290) for carbon storage in the island forests followed by factor 2 (Beta value of 0.019). To gain insight into the driving factors for carbon storage in the island forests over soil physical-chemical properties, the correlations of moisture, pH, salinity, total nitrogen, total phosphorus, total organic carbon, total carbon, C/N ratios in the island soil was calculated. Factor 1 was characterized by high loadings of the parameters total nitrogen (0.842), total organic carbon (0.899), total carbon (0.990). Factor 2 was driven by total phosphorus (0.931), C/N ratios (-0.925). Factor 3 was characterized by moisture (0.694), pH (0.744), salinity (-0.666). A multiple linear regression model showed factor 3 as most important factor (Beta value of-0.694) for the carbon storage in the island forests followed by factor 1 (Beta value of-0.192) and factor 2 (Beta value of-0.106). We conclude that soil texture, pH, moisture and salinity in the island soil were primary controlling factors on the average carbon stock of the forests on the Beichangshan island.