The basic procedure follows closely the Bradley chapter from Methods in Enzymology volume 225: Guide to techniques in mouse development (pp. 855-877, chapter 51). This provides a basic guide but we have made several modifications to it. In particular we pick colonies in media into media. Before starting you must have a dedicated octapipetteman and Inotech octapette aspirator--if you don't then use one of the older 24-well techniques. Once trained it takes approximately one hour to pick a 96-well plate, so if you're going to do 5 plates plan accordingly.

-After 7-10 days of selection the colonies should be visible to the naked eye and most of the debris should have cleared away (feed every other day until ready to pick and discontinue any negative selection).

-Coat a 96-well plate as usual and fill with 180 µL of medium + G418. Warm up in incubator for at least 10-15 min.

-Remove 100 mm dish and fresh 96-well from incubator. Draw a convenient grid on the bottom of the 100 mm dishes (I prefer to pick a different 100 mm dish for each 96-well and rotate between the dishes and plates. That way no pair of plates stays out for an excessive amount of time).

-Use a P10 set to 10 µL and clean unbeveled yellow tips to pick colonies (I use pulled glass pipettes and a mouth pipetor, but everyone else has used yellow tips and it is much faster to learn and become proficient at it this way).

-Depress the plunger on the P10 and nudge the colony to dislodge it. Then rotate the P10 to a vertical position and release the plunger upward to suck up the colony. The colony should be visible in the tip. (this takes some practice but soon becomes second nature).

-Move to the 96-well and position the tip at the junction of the bottom of a well and its side (see diagram). Draw up several times to break the colony up into 3-5 pieces. Leave a bubble in the well to indicate each well that has received a colony.

-Examine the 96-well to see how you did. Usually this takes a while to become proficient, so you won't always have to check. If the colony is not well broken up it will be OK–3-5 pieces is a preferred goal.

-Continue picking for a couple of rows, using a clean tip each time. Then rotate to another pair of plates to give the current set a break.

-Once finished return the 96-wells to the incubator: ignore them the next day, except to check how they are settling down. Assuming this looks all-right you can now discard the primary selection plates.