This paper introduces the *Expected Transmission Count* (ETX) metric, which is designed to find high-throughput paths in multi-hop wireless networks. ETX minimizes the expected total number of packet transmissions required to deliver a packet to its destination, taking into account link loss ratios, asymmetry in loss ratios between links, and interference among successive links. In contrast to the commonly used minimum hop-count metric, which chooses paths based on hop count regardless of throughput, ETX provides a more accurate and efficient routing solution. The paper describes the design and implementation of ETX for the DSDV and DSR routing protocols, along with modifications to these protocols to enable their use with ETX. Experiments conducted on a 29-node 802.11b test-bed demonstrate that ETX significantly improves throughput compared to the minimum hop-count metric, especially for longer paths. The results show that ETX can improve throughput by up to two times over the minimum hop-count metric, highlighting its effectiveness in improving network performance. The paper also explores the performance of minimum hop-count routing, explaining why it often finds routes with lower throughput than the best available paths. It discusses the distribution of path throughputs and link loss ratios, and how these factors affect the performance of routing protocols. The design and implementation details of ETX are provided, including the calculation of link and route ETX values, and the use of link probe packets to measure loss ratios. Finally, the paper evaluates the performance of ETX with both DSDV and DSR, showing that ETX significantly improves initial route selection and overall throughput. The impact of link asymmetry and the accuracy of link measurements are also discussed, providing insights into the strengths and limitations of the ETX metric.This paper introduces the *Expected Transmission Count* (ETX) metric, which is designed to find high-throughput paths in multi-hop wireless networks. ETX minimizes the expected total number of packet transmissions required to deliver a packet to its destination, taking into account link loss ratios, asymmetry in loss ratios between links, and interference among successive links. In contrast to the commonly used minimum hop-count metric, which chooses paths based on hop count regardless of throughput, ETX provides a more accurate and efficient routing solution. The paper describes the design and implementation of ETX for the DSDV and DSR routing protocols, along with modifications to these protocols to enable their use with ETX. Experiments conducted on a 29-node 802.11b test-bed demonstrate that ETX significantly improves throughput compared to the minimum hop-count metric, especially for longer paths. The results show that ETX can improve throughput by up to two times over the minimum hop-count metric, highlighting its effectiveness in improving network performance. The paper also explores the performance of minimum hop-count routing, explaining why it often finds routes with lower throughput than the best available paths. It discusses the distribution of path throughputs and link loss ratios, and how these factors affect the performance of routing protocols. The design and implementation details of ETX are provided, including the calculation of link and route ETX values, and the use of link probe packets to measure loss ratios. Finally, the paper evaluates the performance of ETX with both DSDV and DSR, showing that ETX significantly improves initial route selection and overall throughput. The impact of link asymmetry and the accuracy of link measurements are also discussed, providing insights into the strengths and limitations of the ETX metric.