Landscape pattern is the outcome of comprehensive effects of natural and human factors on geographical space, and has a significant impact on the structure, function, and process of ecosystems. An archipelago is an aggregation of multiple neighboring islands, and there are significant differences in the basic features such as areas, shapes, and terrain conditions between and within the islands in an archipelago. Meanwhile, human activities, including urban and rural construction, plantations, and transportation, as well as the vulnerability of island ecosystems, are spatially heterogeneous on an archipelago, which make the landscape types more diversified, resulting in more differential landscape patterns and complex ecological effects. In this study, five northern islands of the Miaodao Archipelago were used as study areas, which are the typical islands in North China, located north of the Shandong Peninsula, at the juncture of the Yellow and Bohai Seas. The driving factors of landscape pattern here are highly complicated. Based on field investigations, 3S technology methods were adopted. Patch density (PD), edge density (ED), mean shape index (MSI), and human interference index (HII) were used as the landscape pattern indices. The landscape patterns were analyzed in landscape, island, and grid scales, respectively. Then, the relationships between landscape pattern and ecological effect indicators, including net primary productivity (NPP), plant diversity, and soil properties, were examined. The results revealed that the landscape patterns showed obvious spatial heterogeneity in different scales. In the landscape scale, coniferous forest, broad-leaf forest, and grassland constituted vegetation landscape, which occupied the largest proportion of island landscape, and their PD, ED, and MSI were generally higher. Construction land also had a larger area and its PD was higher but its MSI was minimal; the area of bare land was smaller than those of the above-mentioned types with median landscape indices; the areas of the other landscape types were < 100 hm². In the island scale, both PD and HII increased with increases in island area, population, and GDP. In the grid scale, PD, ED, and MSI had significant positive relationships with altitude, whereas HII had significant negative relationships with altitude and slope. Different landscape pattern indices were also related to each other, and there were significant positive correlations among PD, ED, and MSI. HII had significant negative correlations with PD and ED. Island area, terrain, and human activities were the fundamental, limiting, and direct driving factors of the landscape patterns, respectively. The ecological effects of landscape pattern were closely related to the scale selected. In the landscape scale, all the ecological effect indicators showed significant differences in different landscape types. In the island scale, the responses of ecological effect indicators to landscape pattern were insensitive. In the grid scale, ecological effect indicators and landscape pattern indices had significant simple and partial correlations, yet the results of simple and partial correlations showed differences. NPP and soil moisture content were mainly influenced by landscape types and vegetation status. Biodiversity and soil fertility were influenced by landscape types, and the fragmentation and edge effect of landscape pattern. Increased intensity of human activity has led to a decrease in each ecological effect indicator. Construction scale control, optimization of landscape allocation, and improvement of utilization methods are important measures to maintain the stability of island ecosystems.