Genetic variability and relationships among six native French cattle breeds (Abondance, Tarentaise, Villard de Lans, Montbéliarde, Limousin, and Charolais) and one foreign breed (Holstein) were investigated using 23 microsatellite markers. These breeds were also compared with four Swiss breeds genotyped in a previously published study. Interestingly, the French alpine breeds have smaller population sizes but showed higher genetic variability than the larger Holstein breed. Neighbor-joining trees and PCA (principal components analysis) showed that alpine breeds tend to cluster together. Abondance and Tarentaise breeds were closely related, whereas the Holstein was highly differentiated from all breeds analyzed. Two different assignment tests for determining the breed of origin of individuals were compared: "direct" and "exclusion-simulation" approaches. The exclusion-simulation significance test correctly assigns fewer individuals than the direct approach but provides a confidence level (e.g., P < 0.01) for each individual being assigned. Accurate assignment with high statistical confidence is required for animal traceability. Unfortunately, the accuracy of assignment greatly decreases as the threshold level of confidence of assignment increases (e.g., from P < 0.05 to P < 0.001). Assignment accuracy also greatly declines as the level of population differentiation decreases below the level often found between related breeds (e.g., F(ST) < 0.1).