In this article, three different methods (thermal camera, attached thermocouple and inserted thermocouple) are used to measure sewing needle temperatures on a lockstitch machine. The experiments were conducted at machine speeds from 16 stitches/sec to 66 stitches/sec and for a maximum time of 60 seconds. This research is useful for predicting the real temperature of needles at different sewing speeds and for comparing methods of measurement. The inserted thermocouple method showed the lowest standard deviation and highly reproducible results. The results for all the measurement techniques are compared at different sewing speeds and will be beneficial for researchers who have made theoretical models for sewing needle temperatures and need to verify them using experimental results. The information in this article will also be useful for thread lubricant producers who use thermal cameras for needle temperature measurement. All of our experiments indicate that the needle temperature is always higher when thread is used, which is contradictory to the results of some researchers, who have found that needle temperature is lower when thread is used.