Despite much optimistic language on commercial websites, little data is available on actual performance of hydrokinetic turbines. This paper summarises the findings of a series of tests on several Darrieus type cross flow hydrokinetic turbines (HKTs). Although this type of hydrokinetic turbine (HKT) has some advantages over axial flow turbines, fixed pitch Darrieus HKTs also have some drawbacks, including inability to self-start under load, low efficiency and shaking. Variable pitch has been suggested to increase starting torque and efficiency, ducts to increase power output and helical blades to produce smooth torque. To assess each of these modifications, tests were conducted in Australia and Canada on HKTs with fixed and variable pitch straight blades, fixed helical blades, with and without a slatted diffuser, by mounting each turbine in front of a barge and motoring through still water at speeds ranging from less than 1 m/s up to 5 m/s. The diffuser increased the power output by a factor of 3 in one configuration but considerably less in others. A reason for this finding is suggested. The maximum coefficient of performance Cp of the fixed pitch straight blade and helical turbines without a diffuser ranged from about 0.25 at 1.5 m/s down to less than 0.1 at 5 m/s, while Cp for those with a diffuser ranged from about 0.45 down to about 0.3. Fixed blade turbines, both straight and helical, exhibited low starting torque, while variable pitch turbines started easily. Considerable differences in Cp were observed for the same turbine configuration at different speeds. The turbine with fixed pitch, straight blades was found to shake violently due to cyclical hydrodynamic forces on blades, while the helical and variable pitch turbines did not shake excessively. These findings suggest that variable pitch cross flow HKTs should be further investigated. (C) 2011 Elsevier Ltd. All rights reserved.