Harvesting timber on steep terrain can be both expensive and have higher safety risks. Using ground-based machines on steep terrain has the potential to decrease harvest costs and improve safety through the mechanization of manual forest tasks such as tree felling and choker-setting. To better understand the true range of slopes on which forest machines are operating, a digital accelerometer was attached to 22 forest machines to capture real-time measurements of machine slope. A total of 18 studies were completed in commercial pine plantation operations in New Zealand, and two carried out in Austria and Norway respectively. Site selection deliberately focused on ground-based operations that were working on steep terrain. Machine slope, as measured on the carrier base, was captured using a digital accelerometer at a frequency of 2 Hz, and a GPS unit to capture location data simultaneously. Terrain slope was calculated using two different methods in ArcGIS®; one was based on a triangular irregular network (TIN) file and one was based on a raster file. The evaluated machines were grouped into one of four machine types; felling (n = 4), shovelling (n = 5), skidder (n = 9) or “European” (n = 4). A characterization of the European machines studied was that they were purpose built steep terrain machines, whereas the New Zealand felling and shovelling machines were modified excavators. The machines were analyzed and compared with respect to machine slope (actual) and terrain slope (predicted) based on a digital terrain map and the machines location. All machines studied operated on slopes that exceed the New Zealand Approved Code of Practice guideline of 17 and 22° slope for wheeled and tracked machines respectively. New Zealand-based machines were shown to exceed the guidelines for terrain slope much more frequently, and by a greater margin, than European-based machines. Linear regression showed that the relationship between machine slope and terrain slope was poor for all machines, indicating that terrain slope is a poor predictor for managing machine slope.