Most of the perennial ryegrass cultivars are synthetics composed of 4-20 or more parents. Using component genotypes with a good specific combining ability (SCA) in a synthetic variety may reduce the number of parents and exploit heterosis without inbreeding depression. Determining SCA by carrying out and testing pair crosses is laborious. Is it worthwhile to do all this work? We carried out a diallel cross with 10 genotypes of tetraploid perennial ryegrass. The F1 seeds were multiplied to F1(2) seeds. We determined the dry matter yield of the F1(2) populations in a field trial for 2 years. The results showed that the genotypes with the best general combining ability produced the best F1(2) progenies. The best 3 F1(2) progenies were used to build a synthetic variety that is performing very well in official variety list trials. This example shows that the use of pair crosses in synthetic variety development of tetraploid perennial ryegrass varieties is valuable but not necessarily superior to other less laborious methods like polycross or topcross where GCA is determined.