The Carrington Research Extension Center conducts annual soybean and field pea trials to compare commercial and experimental inoculants of N2-fixing bacteria to an uninoculated treatment and one or more levels of N fertilizer.  Inoculants are supplied by manufacturers and distributors and include peat and liquid (seed applied) and granular (in-furrow) formulations, as well as pre-inoculated seed.  Root systems are visually evaluated for nodulation during the growing season and data is collected on grain yield and quality.  Treatment effects are sometimes quite small and sometimes relatively large.  In 2003, the yield of the best soybean inoculant treatment was 37% higher than the uninoculated check.  This difference was not statistically significant at the 5% level of probability.  It seems odd that a 37% difference is not significant.  However, nodulation and the nitrogen fixation process are very sensitive to environmental conditions.  This is frequently evident among the different replicates of a given treatment or even in adjacent plants within a row.  All too often, one plant may be well-nodulated and an adjacent plant will be sparsely nodulated.  For this reason, 6-10 root systems are examined to determine the nodulation score.  These differences may carry over to yield data, especially on a low-N field which does not have the appropriate bacteria established in the soil.  Every effort is made to eliminate data from plots affected by non-treatment factors (e.g. deer predation).  However, other sources of variability are not as obvious.