Relationships between soil type and ovine and caprine paratuberculosis in the Avila region (central Spain) were evaluated using data from a cross-sectional study of the most-important diseases of small ruminants in this Spanish region between 1996 and 1997. Questionnaire data from 61 herds (38 ovine and 23 caprine) and 1451 serum samples (1041 ovine and 410 caprine) were used. Herd paratuberculosis (herds were scored as positive to paratuberculosis if any of the serum samples was positive in an agar-gel immunodifussion) was the outcome of interest, whereas soil type in the municipality where farms were located was the predictor variable. Other variables related to soil and soil usage, and herd size, replacement, main food production and animal species were also introduced into the multivariable logistic regression. The final model contained only two independent variables: the predictor variable soil type (coded as two dummy variables ST-1 and ST-2) and herd size (dichotomized at the highest quartile). The estimated Odds Ratios were 25.9 (95% CI: 1.6, 411) for ST-1 (entisols as soil type) and 3.5 (95% CI: 0.3, 45) for ST-2 (inceptisols as soil type).