Assimilating accurate behavioral events over a significant period of recorded time can be time consuming. If an automatic device could accurately record duration and frequency for a given behavioral event, this would provide a useful tool for ethology. Eleven gilts were used to compare two methods of recording drinking behavior in pigs; Observer software (OBS) and water meter Hobos (WMHOBO) affixed onto the water line. The method of observation affected (P = 0.0008) the duration and visits (P = 0.0048) at the nipple waterer. The adjusted R 2 value was 0.5633 and 0.6871 for duration and visits to the nipple waterer, respectively. In conclusion, the relationship between methods used to predict time spent at the nipple waterer and number of visits was weak. In addition, OBS method underestimated the number of visits to the nipple waterer and overestimated the total duration of drinking behavior for the nursery pig compared to the WMHOBO. As the relationship between methods was weak, it cannot be concluded that one method is as accurate as the other.