England and Wales Court of Appeal, September 3 2004. Read the full document [PDF]. Excerpt:

In my judgment, the employment tribunal erred in law, as it failed to adopt the approach to disparate adverse impact laid down in Seymour-Smith. It should have taken the statistics for the entire workforce, to which the unfair dismissal and redundancy pay requirement of being under 65 applied. It should then have primarily compared the respective proportions of men and women who could satisfy that requirement. It should not have defined and distorted the relevant pool by excluding the "figures" relied on by the Secretary of State as not relevant for it to consider and by referring only to those who were disadvantaged by the disputed upper age limit requirement. In brief, the employment tribunal erred in regarding as irrelevant to its consideration the figures relied on by the Secretary of State relating to those in the workforce who could comply with the age requirement. If the correct approach is taken, the statistics in evidence clearly establish that the difference in the working population between the proportion of men aged under 65 who can comply and the proportion of women aged under 65 who can comply is very small indeed.

Reported in JURIST's Paper Chase here.