The accuracy and reliability of abundance estimates of the infaunal propellerclam, Cyrtodaria siliqua, derived from their burrow openings was tested over a three-year period after experimental hydraulic clam dredging. Abundances of C. siliqua were obtained from hydraulic grab samples at the same locations as photographs taken to count their burrows. Correlations between abundances of clams and burrows in dredged and nondredged areas were monitored over a three-year period. All correlations between burrows and clams were positive in nondredged areas and three of six correlations were significant. There was a natural trend over time of decreasing numbers of visible burrow openings, whereas abundances of the architect of these burrows, C. siliqua, remained relatively constant. Possible explanations for this phenomenon are discussed. All postdredging correlations in the two dredging boxes were negative and nonsignificant. It is concluded that, in this habitat, counts of burrow openings in photographs are not reliable surrogates of abundances of C. siliqua in grab samples. Monitoring recovery in these populations will require traditional grab sampling.