In mating systems with sperm competition, paternity is frequently established with modern DNA techniques. These methods are often expensive and cumbersome, and can be especially difficult for highly fecund species. An additional objective of many paternity studies is to discover the relationship between sperm number and paternity. We present here a competitive polymerase chain reaction (PCR) protocol, coupled with the use of an automated sequencer, that has two functions: (i) to measure directly relative sperm output of males in sperm competition; and (ii) to estimate paternity distributions of large numbers of offspring simultaneously. Our technique was calibrated using a microsatellite locus of the bluehead wrasse, Thalassoma bifasciatum, with the result that product ratio after competitive PCR accurately reflected the initial template proportions of known mixtures of DNA. When we applied our technique to multiple larvae of separate mating events we found that paternity distributions estimated with the competitive PCR technique closely matched the estimates derived from the traditional method of pooling paternity data from individual larvae. Finally, we compared paternity of these spawns with relative sperm contribution estimates. This comparison suggests that ejaculate size alone does not predict a male’s proportion of paternity within the group.