Winter cereal production in many areas of eastern Canada is limited by a lack of consistent winter survival. A cold acclimated crown freezing test has been adopted to circumvent high variability associated with field tests. Relationships between field survival and crown freezing tolerance (LT50, lethal temperature for 50% of seedlings) of artificially acclimated wheat (Triticura aestivum L.) and barley (Hordeum vulgare L.) were investigated. Field survival was evaluated in 1989 and 1990 at several sites in southern Ontario with a field survival index (FSI). In separate experiments, the effect of rooting medium (calcined clay or soil), cold acclimation regime (constant at 2°C and 12‐h photoperiod or decreasing temperature and photoperiod), and developmental stage (two‐leaf or five‐leaf) on crown freezing tolerance were assessed with six genotypes of wheat and barley. The ability of the crown freezing test to differentiate 11 genotypes each of wheat and barley with from two to 20 replicates was investigated. The correlation between FSI and LT50 was strong, ranging from −0.66 to −0.96 and was not affected by rooting media, developmental stage, or cold acclimation regime. Increasing the number of replicates did not affect the relationship between FSI and LT50, but did increase precision. Eight replications were required to detect differences of 1°C among genotypes. A I°C difference corresponded to a field survival difference of 7.1% for barley and 4.6% for wheat. The range in field survival was narrow at 19.2% for barley and 22.6% for wheat. This implies that, under eastern Canadianc onditions, eight replicates of a crown freezing test could broadly categorize genotypes according to winterhardiness, but would not be sufficient to detect small differences. However, where freezing tests are used to assist in the identification of genotypes with higher levels of winterhardiness, a wide range of protocols can be used.