When selecting pH meter go for the one with replaceable electrode. Depending on the make and the way electrode is used life time is usually in the range of 6-9 months, up to a year at best (that's in the case of the most popular gel type, flowing electrodes can survive much longer - but are difficult to handle). See
www.ph-meter.info for details.
Remember to use correct solution for storing and to calibrate the electrode before use, these things are often overlooked.
I must agree with Mr. pH on the replaceable electrode & calibration. The most common failure mechanism I've dealt with is "dry" reference electrode. The KCL stops flowing either because it finally evaporated out (bad storage conditions) or crystalized in the frit and no amount of soaking in warm (and maybe a touch of hypoallergenic dish soap) buffer 7 solution will get it flowing again. If you can afford it, look for a meter that allows the use of different brands of electrodes, possibly even external references. The isFET electrode meters with stainless steel electrode tips are pretty rugged but a bit on the expensive side.
Somebody mentioned temperature compensation. Many "cheap" meters may not have this feature. But if you have a good glass themometer, you can do it yourself. The equation (for glass electrodes) is 0.003pH/deltaC/deltapH. deltaPH is the deviation away from 7 (so no compensation at pH 7). deltaC is 1/10th the temperature difference from 25C (ie [25-T]/10). This compensation is added to the meter's result. Or, you can Google for "pH temperature compensation" and you will find several application nodes which include a nifty little chart to look it up. As you will see, unless you have both a large temperature differential and a large pH differential, the compenstaion is pretty small.
And most every meter I've ever used has some "drift". I've found I get better results if I warm or chill the calibration buffers to something close to the target temperature. This works best with a meter that does ATC and one where you can enter the calibration buffer's pH as some shift quite a bit when you deviate from 25C. All other things being equal, you can leave your buffers at 25C but definately calibrate at the start of a session, and again a few times during your work session. But clean the probe first!