A series of swirl-flow tests was conducted in which all of the heat was generated in twisted-tape swirl generators. This is in contrast to past ORNL swirl-flow tests with twisted tapes, in which ∼99 percent of the heat was generated in the metallic tube wall. In the present study, water from a constant-head tank flowed by gravity at 5 to 8 fps through a vertical 0.27-in.-ID glass tube ∼13 in. long, in which was located a resistance-heated, 16-mil-thick A-nickel tape. Tape-twist ratios were varied from 2.7 to ∞ inside tube diameters/180-deg twist, inlet water temperatures from 63 to 173 F, and heat fluxes from 0.21 × 106 to 1.20 × 106 Btu/hr·ft2. The water head above the top of the tube was held at 30.7 in. In all cases, the critical wall superheat increased with decrease of tape-twist ratio, whereas the critical heat fluxes for the twisted tapes fell between 93 percent and 122 percent of those for flat tapes, maximizing in all cases at a tape-twist ratio of 7 to 10. It is postulated that the deleterious effect of centripetal acceleration with this geometry, which tends to hold the vapor on the heated surface, is compensated in the swirl-flow entrance region by inertial impingement of the liquid onto the tape surface, and along the remainder of the length by a double-vortex secondary flow pattern in the plane normal to the tube wall. The power density of a swirl-flow tube assembly may therefore be significantly increased by generating heat in the twisted tape as well as in the tube wall.