Abstract: Increasing saliva flow to buffer the stomach has been hypothesized as a basis for crib-biting in horses. Saliva amounts in seven cribbing and seven noncribbing (control) horses were compared either pre- and post-cribbing or at pre- and post-5-minute intervals for controls. A pre-weighed cellulose sponge was used to collect saliva at the exit of the submandibular gland for 30 seconds, then reweighed. Data were analyzed as repeated measures. Mean saliva weight overall was similar between cribbing and control horses (1.2 and 1.5 g, respectively, SE = 0.2). However, mean saliva weight for pre- and post-samples (1.5 and 1.2 g, respectively, SE = 0.06) for all horses was significantly lower (P < .05) in the post-sample, indicating a drying effect of the sponge. Because of a strong tendency (P < .06) for a treatment-by-sampling time interaction, data were analyzed by sampling time and cribbing status. Mean saliva weights in the pre-sample were 0.43 g higher (P < .03) in control than cribbing horses. Control horses showed a 0.38 g decrease (P < .01) in saliva weight between pre- and post-samples, which was not evident in cribbing horses. To determine whether cribbing offset the saliva decrease seen in control horses, nine cribbing horses were sampled as before but prevented from cribbing between samples. A similar reduction (0.39 g, P < .01) in saliva weights between samples with cribbing allowed versus cribbing prevented was seen in these horses as was seen in control horses in the initial study. Because cribbing does produce saliva, gastrointestinal irritation could be a motivating cause for cribbing.