This study was concerned with the following two major hypotheses within the setting of subsistence level farming in Bourkina Fasso : the introduction of an animal traction (AT) program has a positive effect on the productivity per unit land area, and the AT program, known as a labor-saving technology, is also an innovative technology. The study used a functional analysis. From the analysis of results the author concludes: 1. In the case of corn and peanuts, labor, capital and animal traction variables are statistically significant and positive. In the case of millet production, the labor variable is only significant, while in the case of sorghum production the capital variable is most significant, 2. The results show some substitutability between millet and sorghum crop productions, 3. The translog production function estimates for millet suggest several implications such as there exist different levels of technology use between AT and non-AT farms, and there exist not only different shapes of production functions but also different input factor orientations between AT and non-AT farms, 4. Within the context of translog production function in the case of millet production, the results indicate that the production process of non-AT farms seems to be heavily dependent upon labor input and that of AT farms relies on labor, capital and animal power input, and 5. The study draws an inference that in the case of millet production, an AT program may be considered an innovative technology.