The present article aims to compare electromyographic (EMG) activity of the knee extensors during traditional resistance training (TRT) and no load resistance training with or without visual feedback (NL-VF and NL-NF). Sixteen healthy men (age: 25.2 ± 3.6) volunteered to participate in the study. Participants visited the laboratory on three occasions involving: (1) a 10 repetition maximum test (10 RM test), (2) familiarization and (3) performance of knee extensions using TRT, NL-VF and NL-NF in a random order, with 10 min of rest between them. TRT involved the performance of a set to momentary muscle failure using the 10 RM load. NL-NF involved the performance of 10 repetitions with no external load, but with the intention to maximally contract the muscles during the whole set. NL-VF involved the same procedure as NL-NF, but a monitor was positioned in front of the participants to provide visual feedback on the EMG activity. Peak and mean EMG activity were evaluated on the vastus medialis (VM), vastus lateralis (VL) and rectus femoris (RF). Results: there were no significant differences in VM and VL peak EMG activity among different situations. There was a significant difference for peak EMG activity for RF, where TRT resulted in higher values than NL-VF and NL-NF (p < 0.05). Higher values of mean EMG activity were found for VM, VL and RF during TRT in comparison with both NL-VF and NL-NF. Conclusions: resistance training with no external load produced high levels of peak muscle activation, independent of visual feedback, but mean activation was higher during TRT. These results suggest that training with no external load might be used as a strategy for stimulating the knee extensors when there is limited access to specialized equipment. Although the clinical applications of no load resistance training are promising, it is important to perform long-term studies to test if these acute results will reflect in muscle morphological and functional changes.