Farmers often have to comply with several sets of animal welfare regulations, since private standards have been developed in addition to legislation. Using an epidemiological approach, we analysed protocols from animal welfare inspections carried out in Swedish dairy herds by the County Administrative Board (CAB; official control of legislation) and by the dairy company Arla Foods (private control of Arlagården standard) during 2010–2013 in the county of Västra Götaland. CAB and Arla inspections were not carried out simultaneously. We aimed to identify common non-compliances, quantify risk factors of non-compliance, and investigate if non-compliances were based on animal-, resource-, or management-based requirements, as well as determining the time period allowed for achieving compliance. Non-compliance was found in 58% of CAB cases, and 51% of Arla cases (each case comprising a sequence of one or several inspections). Dirty dairy cattle was one of the most frequent non-compliances in both control systems. However, the differences in control results were large, suggesting a difference in focus between the two systems. Tie-stall housing and winter season (Dec–Feb) were common risk factors for non-compliance, and overall organic farms had a lower predicted number of non-compliances compared to conventional farms. The presence of both similarities and differences between the systems underlines the need for transparency, predictability, and clarity of inspections.