A colleague of mine is fond of remarking that positional cloners don't make good cell biol-ogists. While the talents for molecular genealogy and cartography may not match those for elucidating protein function, it is nonetheless the gene hunters who most profoundly confront the task ofanalysing pro-teins that show up with few clues about them. Yet it seems likely that the formidable challenge in deciphering the function of a single protein identified by positional cloning is just the tranquil prelude to the cacophony that attends the analysis of the 80,000 or more proteins to be delivered by the human genome project. And if the challenge in understanding single proteins like those implicated in human disease is daunting, new approaches to function will be required as the universe of sequences expands.