Epidemiological associations, expressed as odds ratios between variables obtained from dairy cow records and sole ulcer occurrence at claw trimmings were estimated with logistic regression analysis on data from 2204 and 1124 cows in lactation 1 and lactations 2 to 9, respectively. Sole ulcer in one foot or more than one foot occurred in 20.0 and 29.7% of cows in lactation 1 and in 23.5 and 24.7% of cows in lactations 2 to 9. The analysis revealed several complicated interactions. Trimming or calving in summer to fall was strongly associated with sole ulcer. Trimming later than 1 to 2 mo after calving was positively associated with sole ulcer depending on milk yield, body weight, or season of calving. If lactation 1 cows were treated for disease (limb, metabolic, digestive, or severe reproductive disorders), sole ulcer in more than one foot occurred earlier in lactation. Milk yield in early lactation and body weight were positively associated with sole ulcer. Heel erosion was associated with sole ulcer depending on season of trimming and parity. Presence of sole ulcer but not heel erosion in a lactation increased the risk of sole ulcer in the subsequent lactation. The interval between trimmings appeared to be important.