The majority of dairy cows are dehorned or genetically hornless. It is argued that this reduces the risk of injurious and stressful social conflicts. On the other hand, in horned herds, management and housing would need to be better adapted to the cows, e.g. by avoiding changes in herd structure or by providing more space and resources, which would contribute to reducing stress. The aim of this cross-sectional study was to explore the effect of horn status on faecal cortisol metabolites (FCM), reflecting stress levels over the past 9-15h, while taking into account compliance with management and housing recommendations to reduce social conflicts. In 25 loose housed dairy herds (12 horned, 13 hornless), 3-4 faecal samples were collected from 674 cows at intervals of 5-15 d (n=2,625 samples). FCMs were analysed by enzyme immunoassay. A total of 15 housing factors and 7 management factors relating to recommendations for keeping horned cattle were recorded. Each factor was categorised as compliant or (partly) non-compliant, and percentages of compliance for ‘housing’ and ‘management’ were calculated. To avoid collinearity, two separate linear mixed models were calculated with the fixed effects of horn status, compliance with 1. housing and 2. management, and the respective interaction with horn status. Cow was always nested in herd as a random effect; days in milk and time of sampling were included as covariates. In general, there was large variation in compliance with recommendations for housing (23-96%) and management (14-79%), though farms with horned herds complied with higher percentages on average. FCM levels did not significantly differ between horned and hornless herds. However, contrary to expectations, no associations with compliance were found. More knowledge is needed about the extent to which FCM levels reflect social stress and whether individual housing or management factors may be particularly important in reducing stress.