Wolves (Canis lupus) have expanded their distribution into areas of the midwest United States that have not had wolves for several decades. With recolonization of wolves into agricultural areas, there is increasing concern of wolf‐livestock conflicts. To assess the risk wolves may pose to livestock, we initiated a 3‐year study investigating the activity patterns, movements, habitat use, visitation to livestock pastures by wolves, and the occurrence of depredation events in an agricultural‐wildland matrix in northwestern Minnesota, USA. From June 1997 to November 1999, we captured 23 wolves, including pups, from 3 packs; we radiocollared 16 of these wolves. We tracked radioed wolves intensively on a 24‐hour basis during the spring, summer, and autumn of 1998 and 1999. We found wolves passed directly through a pasture containing cattle on 28% of the nights of tracking; 58% and 95% of the wolf locations were ≤1km and ≤5 km from a pasture, respectively. Space use of wolves showed that while they visited livestock pastures during the 24‐hour tracking sessions, they apparently were passing through these pastures with cattle and not preying on livestock. When compared to random simulations of movements, wolves appeared to encounter livestock pastures randomly. Thirty percent of random movements passed directly through a pasture; 65% and 95% of random movements were within ≤1km and ≤5 km of a pasture, respectively. Wolves were more active at night than during the day. Wolves avoided pastures during the day and visited pastures at night when depredations were most likely (i.e., human presence was low). Visitation of livestock pastures was not related to any discernible characteristics of the pastures (i.e., pasture size, cattle density, distance to human habitation, percent forest cover, index of deer abundance). However, pastures in which livestock were killed by wolves contained more cattle than pastures without depredations, but in 1998 only. While the risk of wolf predation on livestock was potentially high (wolves were within ≤1 km of a pasture on 58% of nights), few livestock were actually killed. During the 3‐year study, only 8 animals (all young or vulnerable livestock) were depredated by wolves. Maintaining healthy wild prey populations, removing offending wolves that kill livestock, and encouraging effective and proper husbandry practices (e.g., disposal of carcasses) among livestock producers, should allow for the persistence of wolves in northwestern Minnesota, USA, while minimizing their impact to farmers in this agriculture–wildland matrix.