Two well-known hypotheses from the literature on tournaments are that (1) tournaments can filter out common shocks thereby reducing agents' risk exposure; and (2) disincentive effects can arise when a tournament scheme is administered on a group of mixed ability agents. While handicapping and/or the creation of homogeneous groups have been suggested as mechanisms for mitigating disincentive effects, it is often impractical to use handicapping schemes and nearly impossible to create a completely homogeneous labor force. Hence, contract administrators who intend to use tournaments to elicit effort must be able to assess the positive effects of tournaments (eliminate common shocks) against the negative effects (disincentive effects). Using economic experiments, we find evidence of disincentive effects under tournaments, although these effects are not as strong as predicted. Moreover, tournaments can be effective at reducing earnings variability when common shocks are important. These results suggest that the benefits of risk reduction from eliminating common shocks might outweigh the disincentive effects arising from mixed tournaments.