
If nothing is changing and the detector is at Dynamic Equilibrium, the baseline should not drift. So a drifting baseline is indicative of a change taking place.
Most detectors exhibit a small amount of drift, but this should be hardly noticeable during the day. The instruction manual will list the level of drift specified for the instrument by the manufacturer. Anything in excess of this is either a faulty instrument, or some change in the HPLC conditions which is being detected by the detector:
Changing eluent temperature. This may occur when the eluent bottle is topped up, especially if the solvent store is outside and its cold!
Changing Eluent composition. It shouldn't happen, but may occur if a bottle is topped up.
Changing Flow Rate. This can be because of check valve faults, or air, or a leak. If the pressure changes, this is most noticeable with a Refractive Index detector.
Component eluting from the column. Essentially the system, is not at dynamic equilibrium. Ttry flushing with a strong eluent to clear the column. If it is heavily contaminated, backflushing overnight may be the best way, but don't backflush unless you are sure there is no void in the column.
Contaminent washing out of the flow cell
Bonded phase stripping off the support
Detector lamp or optics warming up