The study of spatiotemporal evolution of forest landscape pattern is beneficial for understanding the process of regional forest change. The traditional patch-based metrics suffer from intrinsic limitations, such as they provide a single average value and cannot provide any spatially explicit information. However, the morphological method can visually identify the type of each landscape element. Meanwhile, the quantitative analysis of integrated fragmentation and aggregation helps to further understand the evolution characteristics of forest landscape pattern. Based on the land use data in 1990, 2000, 2010 and 2020, we established a comprehensive research framework focusing on "Spatial Morphology-Fragmentation-Aggregation" to analyze the evolution characteristics of forest landscape patterns in the Guangdong-Hong Kong-Macao Greater Bay Area. The results showed that, (1) in the past 30 years, the forest core area has been dominated, and the forest area has gradually decreased. However, the decline rate decreased from 2010 to 2020, and all types of forest patches recovered. (2) The overall degree of fragmentation was relatively low, and the central fragmentation area showed a trend of first rising and then falling. (3) The aggregation degree results showed that the forest patches within 500 m from the forest boundary had the largest area, accounting for about 30% of the total area of the study area. The area of non-forest patches with a distance of more than 4.5 km has almost doubled, and forest loss in this area should be a concern. This study contributes to a comprehensive understanding of the changing characteristics of forest landscape patterns. These results can provide a scientific basis for forest ecological restoration projects, territorial spatial planning and sustainable urban development in the Guangdong-Hong Kong-Macao Greater Bay Area.