Since the 1980s, multinational corporations (MNCs) have increasingly invested in production and service facilities in developing countries. In the meantime, developing countries adopted preferential policies to attract foreign direct investment (FDI) in the hope that FDI can bring in advanced technology. However, FDI is a double-edged sword, generating positive as well as negative technology spillovers. Extant studies focus on the simple issue of whether the presence of FDI affects domestic firms, and have limited policy implications for developing countries. This paper is intended to fill this research gap. Differing from extant studies, the paper tries to identify the individual channels through which positive and negative FDI technology spillovers take place in different entry modes, and draws practical implications for developing countries in making policies on MNC entry modes and investment priorities. The paper distinguishes between different forms of FDI presence, including tangible assets and intangible assets, exported products and domestically sold products, new products and traditional products, employment of skilled workers and employment of unskilled workers, and takes them as different channels by which FDI technology spillovers take place. Meanwhile, the paper distinguishes between wholly foreign owned enterprise and joint venture, and takes the two entry modes as different organizational settings in which FDI technology spillovers occur. The paper then examines how FDI technology spillovers take place through each of these channels under each of the entry modes, and tests hypotheses against firm-level data from China. The study finds that positive FDI technology spillovers take place through tangible rather than intangible assets, domestically sold rather than exported products, traditional rather than new products, and employment of unskilled rather than skilled workers in joint ventures.