A key challenge in cartel enforcement is identifying collusive agreements. We study two major Nordic procurement cartels that operated in the asphalt paving market. We find evidence that during the cartel period bids were clustered and the winning bid was isolated. We implement two cartel detection methods that exploit variation in the distribution of bids. The method developed by Clark et al. (forthcoming) correctly rejects competitive bidding for the cartel period in both markets. The method suggested by Huber and Imhof (2019) achieves a high prediction rate in one of the markets but not in the market where the cartel had a more modest impact on bid distribution. Our results suggest that statistical screening methods with low data requirements can be useful for competition authorities in detecting collusive agreements.