To investigate the genetic control of rain check (cuticle cracking) in tomato (Lycopersicon esculentum), a full diallel cross including five parents ranging from very resistant to very susceptible was grown in late spring 1994. A randomized complete-block design with four replications was used and the proportion of fruit showing check was measured on all mature fruit from eight plants per replication at three harvests. Analysis of variance indicated significant (P < 0.0001) variation for line, harvest, and line by harvest interaction. The proportion of fruit affected increased with each successive harvest. Reciprocal differences were tested on a by-harvest basis and found to be nonsignificant. Reciprocals were combined and a Hayman's analysis was performed on a by-harvest basis on the means. Additive effects on variance were significant (P < 0.05) for all harvests. Under high environmental stress (harvest 3), dominance effects were negative and significant (P < 0.05). Narrow-sense heritability ranged from 0.54 to 0.67 and increased with increasing environmental stress. General combining ability was significant for all harvests, whereas specific combining ability was significant only for harvest 3 (P < 0.05).