Objective:

Examine thermocouple model uncertainty (reliability + validity).

Design:

First, a 3 × 3 repeated measures design with independent variables electrothermometers and thermocouple model. Second, a 1 × 3 repeated measures design with independent variable subprobe.

Intervention:

Three electrothermometers, 3 thermocouple models, a multi-sensor probe and a mercury thermometer measured a stable water bath.

Main Outcome Measures:

Temperature and absolute temperature differences between thermocouples and a mercury thermometer.

Results:

Thermocouple uncertainty was greater than manufactures’ claims. For all thermocouple models, validity and reliability were better in the Iso-Themex than the Datalogger, but there were no practical differences between models within an electrothermometers. Validity of multi-sensor probes and thermocouples within a probe were not different but were greater than manufacturers’ claims. Reliability of multiprobes and thermocouples within a probe were within manufacturers claims.

Conclusion:

Thermocouple models vary in reliability and validity. Scientists should test and report the uncertainty of their equipment rather than depending on manufactures’ claims.