The author surveys the literature on trade and foreign direct investment--especially wholly-owned subsidiaries of multinational firms and international joint ventures--as channels for technology transfer. He also discusses licensing and other arm's length channels of technology transfer. He concludes: 1) How trade encourages growth depends on whether knowledge spillover is national or international. Spillover is more likely to be national for developing countries than for industrial countries. 2) Local policy often makes pureforeign direct investment infeasible, so foreign firms choose licensing or joint ventures. The jury is still out on whether licensing or joint ventures lead to more learning by local firms. 3) Policies designed to attract foreign direct investment are proliferating. Several plant-level studies have failed to find positive spillover from foreign direct investment to firms competing directly with subsidiaries of multinationals. (However, these studies treat foreign direct investment as exogenous and assume spillover to be horizontal-when it may be vertical.) All such studies do find the subsidiaries of multinationals to be more productive than domestic firms, so foreign direct investment does result in host countries using resources more effectively. 4) Absorptive capacity in the host country is essential for getting significant benefits from foreign direct investment. Without adequate human capital or investments in research and development, spillover fails to materialize. 5) A country's policy on protection of intellectual property rights affects the type of industry it attracts. Firms for which such rights are crucial (such as pharmaceutical firms) are unlikely to invest directly in countries where such protections are weak, or will not invest in manufacturing and research and development activities. Policy on intellectual property rights also influences whether technology transfer comes through licensing, joint ventures, or the establishment of wholly-owned subsidiaries.