Salivary secretory immunoglobulin A (sIgA) concentrations of prospective guide dogs for the blind were determined to clarify whether salivary sIgA is useful in evaluating the potential suitability of guide dogs for the blind. Saliva was collected from 73 prospective guide dogs in the kennel on day 1 (the day of separation from volunteer puppy-raisers), 2, 3, 7 and 14 during the estimation period (at about 1 year old). We selected particularly suitable dogs (superior dogs) and unsuitable dogs (inferior dogs) on the basis of the trainers' estimation. All dogs were divided into two groups, those were acceptable dogs would become the guide dogs and rejected dogs could not become guide dogs. The sIgA concentrations in superior dogs gradually increased from day 1 to 14 and those in inferior dogs remained at low levels. Moreover, the sIgA concentrations on day 14 in the acceptable dogs were significantly higher than those in rejected dogs. The cut-off point of sIgA concentrations on day 14 using an ROC curve was 90 EU/ml, and the specificity of the estimation at this point (70.4%) was higher than that of trainers' estimation (50%). Moreover, parallel testing using both trainers' estimation and sIgA estimation, showed that specificity was further improved (79.5%). The present study showed that sIgA concentration was extremely useful in estimating the adaptive ability for guide dogs for the blind.