On highways, truck platooning may reduce fuel consumption, improve road safety and streamline trucking operations. However, most roads worldwide are two-way, two-lane rural roads, i.e., conditions for which truck platooning should be tested to explore the extent of those advantages. This paper reports findings from a field study undertaken in Northern Norway, testing a platoon of three semi-automated trucks on rural roads with tunnels, mountain passes and adverse geometries. Fleet management and distance data, videos, interviews and conversations between participants were used to assess whether platooning was feasible on such roads. The platooning system was used without interventions through most road conditions, and worked well on flat and wide roads with 90 km/h speed limits. However, it struggled in sharp horizontal curves, where the following trucks would speed up before regaining connection to their preceding truck and then brake abruptly to regain the prescribed distance. Moreover, steep uphills were problematic due to inconsistent gear shifting between the trucks. Seemingly, no fuel savings were achieved, due to excessive following distances and suboptimal speed profiles on crest curves. To obtain further insights into the benefits of truck platooning on rural roads, we suggest redoing the field study with V2V-communication, allowing for shorter following distances, and also performing a manual-driven baseline first.