In esure Insurance Ltd v Direct Line Insurance plc [2008] EWCA Civ 842 (full text here; facts and IPKat analysis here), the Court of Appeal for England and Wales yesterday discussed the issue of whether the European Court of Justice in Case C-106/03 P Vedial v OHIM established a threshold level for the determination of whether two marks are similar to one another.

The Court confirmed that confusion had to be ascertained from the viewpoint of the average consumer in the light of all the relevant factors. There was no threshold condition of similarity. It is for the court or trade mark office to determine what the average consumer would have thought of the two marks and whether they would have confused him. Since the critical issue of confusion of any kind was to be assessed from the viewpoint of the average consumer, little would be gained from the evidence of an expert as to what his own opinion was, when the relevant tribunal was in a position to form its own view.

As Lady Justice Arden said, on the issue of "thresholds":

"49 In a lengthy passage ... the judge considers whether there is a minimum or threshold level of similarity, which is required to be shown. ... The judge's main basis for [finding that a threshold existed] was an inference which he drew from Vedial SA v OHIM, France Distribution Case C-106/03 P of October 2004 [2004] ECR I-9573. In that case, the Court of Justice dismissed an appeal against the decision of the Court of First Instance that two marks both using the name Hubert but one using that name with the prefix "Saint" were not similar. The judge held that this meant that the similarity in respect of the words used in both marks was insufficient to amount to similarity for the purpose of the trade mark directive. But it does not follow that the reason for the Court's conclusion was that there was some minimum level of similarity. In my judgment, it held that there was a failure to show similarity on a global assessment of all the relevant factors. I do not find any threshold condition in the jurisprudence of the Court of Justice cited to us. Moreover I consider that no useful purpose is served by holding that there is some minimum threshold level of similarity that has to be shown. If there is no similarity at all, there is no likelihood of confusion to be considered. If there is some similarity, then the likelihood of confusion has to be considered but it is unnecessary to interpose a need to find a minimum level of similarity.

50 Accordingly, I incline to the view that there is no threshold condition of similarity. ..."