This paper evaluates the effectiveness of remote inspections using a robot in a laboratory experiment. The experiment differs from most human-robot interaction experiments in its direct comparison of manned and robotic operations. 21 participants each performed three manned inspections and three inspections with each of the two remote inspection methods, teleoperated and assisted. The effectiveness was measured based on the number of errors they were able to identify. Teleoperated inspections were found to be less effective than manned, although this difference was not statistical significant. Assisted inspections, implemented as an interactive simulation prototype representing a robot with higher autonomy, had similar effectiveness as manned. Because of the time and high cost required for manned inspections of offshore wind turbines, remote inspection can give a large economic benefit. However, this will only be a viable alternative if the robot system is inexpensive and remote inspections are as effective for identifying errors as manned inspections, which the experiment presented here suggests.